Natural Language Processing (CSEP 517): Language Models, Continued
|
|
- Nathaniel Atkins
- 5 years ago
- Views:
Transcription
1 Natural Language Processing (CSEP 517): Language Models, Continued Noah Smith c 2017 University of Washington nasmith@cs.washington.edu April 3, / 102
2 To-Do List Online quiz: due Sunday Print, sign, and return the academic integrity statement (if you haven t already) Read: Smith (2017); optionally, Jurafsky and Martin (2016), Collins (2011) 2, and Goldberg (2015) 0 4, if you want to know more about neural networks A1 now due April 9 (Sunday) Late policy: four late days 2 / 102
3 Language Models: Definitions V is a finite set of (discrete) symbols ( words or possibly characters); V = V V is the (infinite) set of sequences of symbols from V whose final symbol is p : V R, such that: For any x V, p(x) 0 p(x = x) = 1 x V (I.e., p is a proper probability distribution.) Language modeling: estimate p from examples, x 1:n = x 1, x 2,..., x n. Evaluation on test data x 1:m : perplexity, 2 1 m M i=1 log 2 p( x i) 3 / 102
4 Log-Linear Models: Definitions We define a conditional log-linear model p(y X) as: Y is the set of events/outputs ( for language modeling, V) X is the set of contexts/inputs ( for n-gram language modeling, V n 1 ) φ : X Y R d is a feature vector function w R d are the model parameters p w (Y = y X = x) = exp w φ(x, y) exp w φ(x, y ) y Y 4 / 102
5 Breaking It Down p w (Y = y X = x) = exp w φ(x, y) exp w φ(x, y) y Y 5 / 102
6 Breaking It Down exp w φ(x, y) p w (Y = y X = x) = exp w φ(x, y) y Y linear score w φ(x, y) 6 / 102
7 Breaking It Down exp w φ(x, y) p w (Y = y X = x) = exp w φ(x, y) y Y linear score w φ(x, y) nonnegative exp w φ(x, y) 7 / 102
8 Breaking It Down exp w φ(x, y) p w (Y = y X = x) = exp w φ(x, y) y Y linear score w φ(x, y) nonnegative exp w φ(x, y) normalizer exp w φ(x, y ) = Z w (x) y Y 8 / 102
9 Breaking It Down p w (Y = y X = x) = exp w φ(x, y) exp w φ(x, y) y Y linear score w φ(x, y) Log-linear comes from the fact that: nonnegative exp w φ(x, y) normalizer exp w φ(x, y ) = Z w (x) y Y log p w (Y = y X = x) = w φ(x, y) log Z w (x) }{{} constant in y This is an instance of the family of generalized linear models. 9 / 102
10 The Geometric View Suppose we have instance x, Y = {y 1, y 2, y 3, y 4 }, and there are only two features, φ 1 and φ 2. ϕ 2 (x, y 3 ) (x, y 1 ) (x, y 2 ) (x, y 4 ) ϕ 1 As a simple example, let the two features be binary functions. 10 / 102
11 The Geometric View Suppose we have instance x, Y = {y 1, y 2, y 3, y 4 }, and there are only two features, φ 1 and φ 2. ϕ 2 (x, y 3 ) (x, y 1 ) (x, y 2 ) (x, y 4 ) ϕ 1 w φ = w 1 φ 1 + w 2 φ 2 = 0 11 / 102
12 The Geometric View Suppose we have instance x, Y = {y 1, y 2, y 3, y 4 }, and there are only two features, φ 1 and φ 2. ϕ 2 (x, y 3 ) (x, y 1 ) (x, y 2 ) (x, y 4 ) ϕ 1 distance(w φ = 0, φ 0 ) = w φ 0 w 2 w φ 0 12 / 102
13 The Geometric View Suppose we have instance x, Y = {y 1, y 2, y 3, y 4 }, and there are only two features, φ 1 and φ 2. ϕ 2 (x, y 3 ) (x, y 1 ) (x, y 2 ) (x, y 4 ) ϕ 1 w φ(x, y 1 ) > w φ(x, y 3 ) > w φ(x, y 4 ) > 0 w φ(x, y 2 ) 13 / 102
14 The Geometric View Suppose we have instance x, Y = {y 1, y 2, y 3, y 4 }, and there are only two features, φ 1 and φ 2. ϕ 2 (x, y 3 ) (x, y 1 ) (x, y 2 ) (x, y 4 ) ϕ 1 p w (y 1 x) > p w (y 3 x) > p w (y 4 x) > p w (y 2 x) 14 / 102
15 The Geometric View Suppose we have instance x, Y = {y 1, y 2, y 3, y 4 }, and there are only two features, φ 1 and φ 2. ϕ 2 (x, y 3 ) (x, y 1 ) (x, y 2 ) (x, y 4 ) ϕ 1 15 / 102
16 The Geometric View Suppose we have instance x, Y = {y 1, y 2, y 3, y 4 }, and there are only two features, φ 1 and φ 2. ϕ 2 (x, y 3 ) (x, y 1 ) (x, y 2 ) (x, y 4 ) ϕ 1 p w (y 3 x) > p w (y 1 x) > p w (y 2 x) > p w (y 4 x) 16 / 102
17 The Geometric View Suppose we have instance x, Y = {y 1, y 2, y 3, y 4 }, and there are only two features, φ 1 and φ 2. ϕ 2 (x, y 3 ) (x, y 1 ) (x, y 2 ) (x, y 4 ) ϕ 1 Log-linear parameter estimation tries to choose w so that p w (Y x) matches the empirical distribution, c(x,y ) c(x). 17 / 102
18 Why Build Language Models This Way? Exploit features of histories for sharing of statistical strength and better smoothing (Lau et al., 1993) Condition the whole text on more interesting variables like the gender, age, or political affiliation of the author (Eisenstein et al., 2011) Interpretability! Each feature φ k controls a factor to the probability (e w k ). 1 If w k < 0 then φ k makes the event less likely by a factor of e. w k If wk > 0 then φ k makes the event more likely by a factor of e w k. If wk = 0 then φ k has no effect. 18 / 102
19 Log-Linear n-gram Models p w (X = x) = = l p w (X j = x j X 0:j 1 = x 0:j 1 ) j=1 l j=1 assumption = = l j=1 exp w φ(x 0:j 1, x j ) Z w (x 0:j 1 ) l j 1 exp w φ(x j n+1:j 1, x j ) Z w (x j n+1:j 1 ) exp w φ(h j, x j ) Z w (h j ) 19 / 102
20 Example The man who knew too much many little few. hippopotamus 20 / 102
21 What Features in φ(x j n+1:j 1, X j )? 21 / 102
22 What Features in φ(x j n+1:j 1, X j )? Traditional n-gram features: X j 1 = the X j = man 22 / 102
23 What Features in φ(x j n+1:j 1, X j )? Traditional n-gram features: X j 1 = the X j = man Gappy n-grams: X j 2 = the X j = man 23 / 102
24 What Features in φ(x j n+1:j 1, X j )? Traditional n-gram features: X j 1 = the X j = man Gappy n-grams: X j 2 = the X j = man Spelling features: X j s first character is capitalized 24 / 102
25 What Features in φ(x j n+1:j 1, X j )? Traditional n-gram features: X j 1 = the X j = man Gappy n-grams: X j 2 = the X j = man Spelling features: X j s first character is capitalized Class features: X j is a member of class / 102
26 What Features in φ(x j n+1:j 1, X j )? Traditional n-gram features: X j 1 = the X j = man Gappy n-grams: X j 2 = the X j = man Spelling features: X j s first character is capitalized Class features: X j is a member of class 132 Gazetteer features: X j is listed as a geographic place name 26 / 102
27 What Features in φ(x j n+1:j 1, X j )? Traditional n-gram features: X j 1 = the X j = man Gappy n-grams: X j 2 = the X j = man Spelling features: X j s first character is capitalized Class features: X j is a member of class 132 Gazetteer features: X j is listed as a geographic place name You can define any features you want! Too many features, and your model will overfit Too few (good) features, and your model will not learn 27 / 102
28 What Features in φ(x j n+1:j 1, X j )? Traditional n-gram features: X j 1 = the X j = man Gappy n-grams: X j 2 = the X j = man Spelling features: X j s first character is capitalized Class features: X j is a member of class 132 Gazetteer features: X j is listed as a geographic place name You can define any features you want! Too many features, and your model will overfit Feature selection methods, e.g., ignoring features with very low counts, can help. Too few (good) features, and your model will not learn 28 / 102
29 Feature Engineering Many advances in NLP (not just language modeling) have come from careful design of features. 29 / 102
30 Feature Engineering Many advances in NLP (not just language modeling) have come from careful design of features. Sometimes feature engineering is used pejoratively. 30 / 102
31 Feature Engineering Many advances in NLP (not just language modeling) have come from careful design of features. Sometimes feature engineering is used pejoratively. Some people would rather not spend their time on it! 31 / 102
32 Feature Engineering Many advances in NLP (not just language modeling) have come from careful design of features. Sometimes feature engineering is used pejoratively. Some people would rather not spend their time on it! There is some work on automatically inducing features (Della Pietra et al., 1997). 32 / 102
33 Feature Engineering Many advances in NLP (not just language modeling) have come from careful design of features. Sometimes feature engineering is used pejoratively. Some people would rather not spend their time on it! There is some work on automatically inducing features (Della Pietra et al., 1997). More recent work in neural networks can be seen as discovering features (instead of engineering them). 33 / 102
34 Feature Engineering Many advances in NLP (not just language modeling) have come from careful design of features. Sometimes feature engineering is used pejoratively. Some people would rather not spend their time on it! There is some work on automatically inducing features (Della Pietra et al., 1997). More recent work in neural networks can be seen as discovering features (instead of engineering them). But in much of NLP, there s a strong preference for interpretable features. 34 / 102
35 How to Estimate w? p θ (x) = n-gram l j=1 θ xj h j log-linear n-gram l j 1 exp w φ(h j, x j ) Z w (h j ) Parameters: θ v h w k v V, h (V { }) n 1 k {1,..., d} MLE: ˆθv h = c(hv) c(h) no closed form 35 / 102
36 MLE for w Let training data consist of {(h i, x i )} N i=1. 36 / 102
37 MLE for w Let training data consist of {(h i, x i )} N i=1. Maximum likelihood estimation is: max w R d N log p w (x i h i ) i=1 = max w R d = max w R d N i=1 N log exp w φ(h i, x i ) Z w (h i ) i=1 w φ(h i, x i ) log v V exp w φ(h i, v) } {{ } Z w(h i ) 37 / 102
38 MLE for w Let training data consist of {(h i, x i )} N i=1. Maximum likelihood estimation is: max w R d N log p w (x i h i ) i=1 = max w R d = max w R d This is concave in w. N i=1 N log exp w φ(h i, x i ) Z w (h i ) i=1 w φ(h i, x i ) log v V exp w φ(h i, v) } {{ } Z w(h i ) 38 / 102
39 MLE for w Let training data consist of {(h i, x i )} N i=1. Maximum likelihood estimation is: max w R d N log p w (x i h i ) i=1 = max w R d = max w R d This is concave in w. N i=1 N log exp w φ(h i, x i ) Z w (h i ) i=1 w φ(h i, x i ) log v V Z w (h i ) involves a sum over V terms. exp w φ(h i, v) } {{ } Z w(h i ) 39 / 102
40 MLE for w max w R d N w φ(h i, x i ) log Z w (h i ) }{{} f i (w) i=1 40 / 102
41 MLE for w max w R d N w φ(h i, x i ) log Z w (h i ) }{{} f i (w) i=1 Hope/fear view: for each instance i, increase the score of the correct output x i, score(x i ) = w φ(h i, x i ) decrease the softened max score overall, log v V exp score(v) 41 / 102
42 MLE for w max w R d N w φ(h i, x i ) log Z w (h i ) }{{} f i (w) i=1 Gradient view: w f i = φ(h i, x i ) p }{{} w (v h i ) φ(h i, v) v V observed features }{{} expected features Setting this to zero means getting model s expectations to match empirical observations. 42 / 102
43 MLE for w: Algorithms Batch methods (L-BFGS is popular) Stochastic gradient ascent/descent more common today, especially with special tricks for adapting the step size over time Many specialized methods (e.g., iterative scaling ) 43 / 102
44 Stochastic Gradient Descent Goal: minimize N i=1 f i(w) with respect to w. Input: initial value w, number of epochs T, learning rate α For t {1,..., T }: Choose a random permutation π of {1,..., N}. For i {1,..., N}: w w α w f π(i) Output: w 44 / 102
45 Avoiding Overfitting Maximum likelihood estimation: max w R d N w φ(h i, x i ) log Z w (h i ) i=1 If φ j (h, x) is (almost) always positive, we can always increase the objective (a little bit) by increasing w j toward / 102
46 Avoiding Overfitting Maximum likelihood estimation: max w R d N w φ(h i, x i ) log Z w (h i ) i=1 If φ j (h, x) is (almost) always positive, we can always increase the objective (a little bit) by increasing w j toward +. Standard solution is to add a regularization term: max w R d N i=1 w φ(h i, x i ) log v V where λ > 0 is a hyperparameter and p = 2 or 1. exp w φ(h i, v) λ w p p 46 / 102
47 MLE for w If we had more time, we d study this problem more carefully! Here s what you must remember: There is no closed form; you must use a numerical optimization algorithm like stochastic gradient descent. Log-linear models are powerful but expensive (Z w (h i )). Regularization is very important; we don t actually do MLE. Just like for n-gram models! Only even more so, since log-linear models are even more expressive. 47 / 102
48 Quick Recap Two kinds of language models so far: representation? estimation? think about? n-gram h i is (n 1) previous symbols count and normalize smoothing log-linear featurized representation of h i, x i iterative gradient descent features 48 / 102
49 Neural Network: Definitions Warning: there is no widely accepted standard notation! A feedforward neural network n ν is defined by: A function family that maps parameter values to functions of the form n : R d in R dout ; typically: Non-linear Differentiable with respect to its inputs Assembled through a series of affine transformations and non-linearities, composed together Symbolic/discrete inputs handled through lookups. Parameter values ν Typically a collection of scalars, vectors, and matrices We often assume they are linearized into R D 49 / 102
50 A Couple of Useful Functions softmax : R k R k x 1, x 2,..., x k e x 1 k, j=1 ex j e x 2 k,..., j=1 ex j e x k k j=1 ex j tanh : R [ 1, 1] x ex e x e x + e x Generalized to be elementwise, so that it maps R k [ 1, 1] k. Others include: ReLUs, logistic sigmoids, PReLUs, / 102
51 One Hot Vectors Arbitrarily order the words in V, giving each an index in {1,..., V }. Let e i R V contain all zeros, with the exception of a 1 in position i. This is the one hot vector for the ith word in V. 51 / 102
52 Feedforward Neural Network Language Model (Bengio et al., 2003) Define the n-gram probability as follows: ( p( h 1,..., h n 1 ) = n ν eh1,..., e hn 1 ) = n 1 softmax + M tanh b V j=1 e hj V A j V d d V + W V H u H e h j M T j n 1 + where each e hj R V is a one-hot vector and H is the number of hidden units in the neural network (a hyperparameter ). Parameters ν include: M R V d, which are called embeddings (row vectors), one for every word in V Feedforward NN parameters b R V, A R (n 1) d V, W R V H, u R H, T R (n 1) d H j=1 d H 52 / 102
53 Breaking It Down Look up each of the history words h j, j {1,..., n 1} in M; keep two copies. e hj V e hj V M V d M V d 53 / 102
54 Breaking It Down Look up each of the history words h j, j {1,..., n 1} in M; keep two copies. Rename the embedding for h j as m hj. e hj M = m hj e hj M = m hj 54 / 102
55 Breaking It Down Apply an affine transformation to the second copy of the history-word embeddings (u, T) m hj u H n 1 + j=1 m hj T j d H 55 / 102
56 Breaking It Down Apply an affine transformation to the second copy of the history-word embeddings (u, T) and a tanh nonlinearity. m hj n 1 tanh u + m hj T j j=1 56 / 102
57 Breaking It Down Apply an affine transformation to everything (b, A, W). b V m hj A j n 1 + j=1 + W V H d V n 1 tanh u + m hj T j j=1 57 / 102
58 Breaking It Down Apply a softmax transformation to make the vector sum to one. n 1 softmax b + m hj A j j=1 n 1 + W tanh u + m hj T j j=1 58 / 102
59 Breaking It Down n 1 softmax b + m hj A j j=1 n 1 + W tanh u + m hj T j j=1 Like a log-linear language model with two kinds of features: Concatenation of context-word embeddings vectors m hj tanh-affine transformation of the above New parameters arise from (i) embeddings and (ii) affine transformation inside the nonlinearity. 59 / 102
60 Visualization M u, T tanh W softmax b, A 60 / 102
61 Number of Parameters D = }{{} V d + }{{} V + (n 1)dV + }{{}}{{} V H + }{{} H + (n 1)dH }{{} M b A W u T For Bengio et al. (2003): V (after OOV processing) d {30, 60} H {50, 100} n 1 = 5 So D = 461V parameters, compared to O(V n ) for classical n-gram models. Forcing A = 0 eliminated 300V parameters and performed a bit better, but was slower to converge. If we averaged m hj instead of concatenating, we d get to 221V (this is a variant of continuous bag of words, Mikolov et al., 2013). 61 / 102
62 Why does it work? 62 / 102
63 Why does it work? Historical answer: multiple layers and nonlinearities allow feature combinations a linear model can t get. 63 / 102
64 Why does it work? Historical answer: multiple layers and nonlinearities allow feature combinations a linear model can t get. Suppose we want y = xor(x1, x 2 ); this can t be expressed as a linear function of x 1 and x / 102
65 xor Example x 2 x 1 y Tuples where y = xor(x 1, x 2 ) are red; tuples where y xor(x 1, x 2 ) are blue. 65 / 102
66 Why does it work? Historical answer: multiple layers and nonlinearities allow feature combinations a linear model can t get. Suppose we want y = xor(x 1, x 2 ); this can t be expressed as a linear function of x 1 and x 2. But: z = x 1 x 2 y = x 1 + x 2 2z 66 / 102
67 xor Example (D = 13) Credit: Chris Dyer ( min mean squared error v,a,w,b x 1 {0,1} x 2 {0,1} min v,a,w,b x 1 {0,1} x 2 {0,1} iterations ( ( xor(x 1, x 2 ) v 3 xor(x 1, x 2 ) v 3 ( W 3 2 x 2 ( tanh ) ) 2 + b + a 3 W 3 2 x 2 ) ) 2 + b + a 3 67 / 102
68 Why does it work? Historical answer: multiple layers and nonlinearities allow feature combinations a linear model can t get. Suppose we want y = xor(x 1, x 2 ); this can t be expressed as a linear function of x 1 and x 2. But: z = x 1 x 2 y = x 1 + x 2 2z With high-dimensional inputs, there are a lot of conjunctive features to search through. For log-linear models, Della Pietra et al. (1997) did this, greedily. 68 / 102
69 Why does it work? Historical answer: multiple layers and nonlinearities allow feature combinations a linear model can t get. Suppose we want y = xor(x 1, x 2 ); this can t be expressed as a linear function of x 1 and x 2. But: z = x 1 x 2 y = x 1 + x 2 2z With high-dimensional inputs, there are a lot of conjunctive features to search through. For log-linear models, Della Pietra et al. (1997) did this, greedily. Neural models seem to smoothly explore lots of approximately-conjunctive features. 69 / 102
70 Why does it work? Historical answer: multiple layers and nonlinearities allow feature combinations a linear model can t get. Suppose we want y = xor(x 1, x 2 ); this can t be expressed as a linear function of x 1 and x 2. But: z = x 1 x 2 y = x 1 + x 2 2z With high-dimensional inputs, there are a lot of conjunctive features to search through. For log-linear models, Della Pietra et al. (1997) did this, greedily. Neural models seem to smoothly explore lots of approximately-conjunctive features. Modern answer: representations of words and histories are tuned to the prediction problem. 70 / 102
71 Why does it work? Historical answer: multiple layers and nonlinearities allow feature combinations a linear model can t get. Suppose we want y = xor(x 1, x 2 ); this can t be expressed as a linear function of x 1 and x 2. But: z = x 1 x 2 y = x 1 + x 2 2z With high-dimensional inputs, there are a lot of conjunctive features to search through. For log-linear models, Della Pietra et al. (1997) did this, greedily. Neural models seem to smoothly explore lots of approximately-conjunctive features. Modern answer: representations of words and histories are tuned to the prediction problem. Word embeddings: a powerful idea / 102
72 Important Idea: Words as Vectors The idea of embedding words in R d is much older than neural language models. 72 / 102
73 Important Idea: Words as Vectors The idea of embedding words in R d is much older than neural language models. You should think of this as a generalization of the discrete view of V. 73 / 102
74 Important Idea: Words as Vectors The idea of embedding words in R d is much older than neural language models. You should think of this as a generalization of the discrete view of V. Considerable ongoing research on learning word representations to capture linguistic similarity (Turney and Pantel, 2010); this is known as vector space semantics. 74 / 102
75 Words as Vectors: Example baby cat 75 / 102
76 Words as Vectors: Example baby pig mouse cat 76 / 102
77 Parameter Estimation Bad news for neural language models: Log-likelihood function is not concave. So any perplexity experiment is evaluating the model and an algorithm for estimating it. Calculating log-likelihood and its gradient is very expensive (5 epochs took 3 weeks on 40 CPUs). 77 / 102
78 Parameter Estimation Bad news for neural language models: Log-likelihood function is not concave. So any perplexity experiment is evaluating the model and an algorithm for estimating it. Calculating log-likelihood and its gradient is very expensive (5 epochs took 3 weeks on 40 CPUs). Good news: n ν is differentiable with respect to M (from which its inputs come) and ν (its parameters), so gradient-based methods are available. Essential: the chain rule from calculus (sometimes called backpropagation ) Lots more details in Bengio et al. (2003) and (for NNs more generally) in Goldberg (2015). 78 / 102
79 Next Up More examples of neural language models (in brief): The log-bilinear language model Recurrent neural network language models 79 / 102
80 Log-Bilinear Language Model (Mnih and Hinton, 2007) Define the n-gram probability as follows, for each v V: n 1 ( exp p(v h 1,..., h n 1 ) = j=1 n 1 ( exp v V j=1 m hj A j d d d m hj A j d d d ) + b m v d d ) + b m v d d +c v +c v 80 / 102
81 Log-Bilinear Language Model (Mnih and Hinton, 2007) Define the n-gram probability as follows, for each v V: n 1 ( exp p(v h 1,..., h n 1 ) = j=1 n 1 ( exp v V j=1 m hj A j d d d m hj A j Number of parameters: D = }{{} V d + (n 1)d 2 + d }{{} M A d d d }{{} b ) + b m v d d ) + b m v d d + V }{{} c +c v +c v 81 / 102
82 Log-Bilinear Language Model (Mnih and Hinton, 2007) Define the n-gram probability as follows, for each v V: n 1 ( exp p(v h 1,..., h n 1 ) = j=1 n 1 ( exp v V j=1 m hj A j d d d m hj A j d d d ) + b m v d d ) + b m v d d +c v +c v Number of parameters: D = }{{} V d + (n 1)d 2 + }{{}}{{} d + }{{} V M A b c The predicted word s probability depends on its vector m v, not just on the vectors of the history words. 82 / 102
83 Log-Bilinear Language Model (Mnih and Hinton, 2007) Define the n-gram probability as follows, for each v V: n 1 ( exp p(v h 1,..., h n 1 ) = j=1 n 1 ( exp v V j=1 m hj A j d d d m hj A j d d d ) + b m v d d ) + b m v d d +c v +c v Number of parameters: D = }{{} V d + (n 1)d 2 + }{{}}{{} d + }{{} V M A b c The predicted word s probability depends on its vector m v, not just on the vectors of the history words. Training this model involves a sum over the vocabulary (like log-linear models we saw earlier). 83 / 102
84 Log-Bilinear Language Model (Mnih and Hinton, 2007) Define the n-gram probability as follows, for each v V: n 1 ( exp p(v h 1,..., h n 1 ) = j=1 n 1 ( exp v V j=1 m hj A j d d d m hj A j d d d ) + b m v d d ) + b m v d d +c v +c v Number of parameters: D = }{{} V d + (n 1)d 2 + }{{}}{{} d + }{{} V M A b c The predicted word s probability depends on its vector m v, not just on the vectors of the history words. Training this model involves a sum over the vocabulary (like log-linear models we saw earlier). Later work explored variations to make learning faster. 84 / 102
85 Observations about Neural Language Models (So Far) There s no knowledge built in that the most recent word h n 1 should generally be more informative than earlier ones. This has to be learned. In addition to choosing n, also have to choose dimensionalities like d and H. Parameters of these models are hard to interpret. 85 / 102
86 Observations about Neural Language Models (So Far) There s no knowledge built in that the most recent word h n 1 should generally be more informative than earlier ones. This has to be learned. In addition to choosing n, also have to choose dimensionalities like d and H. Parameters of these models are hard to interpret. Example: l2 -norm of A j and T j in the feedforward model correspond to the importance of history position j. Individual word embeddings can be clustered and dimensions can be analyzed (e.g., Tsvetkov et al., 2015). 86 / 102
87 Observations about Neural Language Models (So Far) There s no knowledge built in that the most recent word h n 1 should generally be more informative than earlier ones. This has to be learned. In addition to choosing n, also have to choose dimensionalities like d and H. Parameters of these models are hard to interpret. Architectures are not intuitive. 87 / 102
88 Observations about Neural Language Models (So Far) There s no knowledge built in that the most recent word h n 1 should generally be more informative than earlier ones. This has to be learned. In addition to choosing n, also have to choose dimensionalities like d and H. Parameters of these models are hard to interpret. Architectures are not intuitive. Still, impressive perplexity gains got people s interest. 88 / 102
89 Recurrent Neural Network Each input element is understood to be an element of a sequence: x 1, x 2,..., x l At each timestep t: The tth input element x t is processed alongside the previous state s t 1 to calculate the new state (s t ). The tth output is a function of the state st. The same functions are applied at each iteration: s t = f recurrent (x t, s t 1 ) y t = f output (s t ) In RNN language models, words and histories are represented as vectors (respectively, x t = e xt and s t ). 89 / 102
90 RNN Language Model The original version, by Mikolov et al. (2010) used a simple RNN architecture along these lines: ( ( ) s t = f recurrent (e xt, s t 1 ) = sigmoid e x t M) A + s t 1 B + c ( ) y t = f output (s t ) = softmax s t U p(v x 1,..., x t 1 ) = [y t ] v Note: this is not an n-gram (Markov) model! 90 / 102
91 Visualization M sigmoid A, B, c U softmax s t s t - 1 x t y t 91 / 102
92 Visualization M A, B, c sigmoid U softmax M A, B, c sigmoid U softmax M A, B, c sigmoid U softmax M A, B, c sigmoid U softmax 92 / 102
93 Improvements to RNN Language Models The simple RNN is known to suffer from two related problems: Vanishing gradients during learning make it hard to propagate error into the distant past. State tends to change a lot on each iteration; the model forgets too much. Some variants: Stacking these functions to make deeper networks. Sundermeyer et al. (2012) use long short-term memories (LSTMs) and Cho et al. (2014) use gated recurrent units (GRUs) to define f recurrent. Mikolov et al. (2014) engineer the linear transformation in the simple RNN for better preservation. Jozefowicz et al. (2015) used randomized search to find even better architectures. 93 / 102
94 Comparison: Probabilistic vs. Connectionist Modeling Probabilistic Connectionist What do we engineer? features, assumptions architectures Theory? as N gets large not really Interpretation of parameters? often easy usually hard 94 / 102
95 Parting Shots 95 / 102
96 Parting Shots I said very little about estimating the parameters. 96 / 102
97 Parting Shots I said very little about estimating the parameters. At present, this requires a lot of engineering. 97 / 102
98 Parting Shots I said very little about estimating the parameters. At present, this requires a lot of engineering. New libraries to help you are coming out all the time. 98 / 102
99 Parting Shots I said very little about estimating the parameters. At present, this requires a lot of engineering. New libraries to help you are coming out all the time. Many of them use GPUs to speed things up. 99 / 102
100 Parting Shots I said very little about estimating the parameters. At present, this requires a lot of engineering. New libraries to help you are coming out all the time. Many of them use GPUs to speed things up. This progression is worth reflecting on: history: represented as: before 1996 (n 1)-gram discrete feature vector embedded vector since 2010 unrestricted embedded 100 / 102
101 References I Yoshua Bengio, Réjean Ducharme, Pascal Vincent, and Christian Jauvin. A neural probabilistic language model. Journal of Machine Learning Research, 3(Feb): , URL Kyunghyun Cho, Bart van Merrienboer, Caglar Gulcehre, Dzmitry Bahdanau, Fethi Bougares, Holger Schwenk, and Yoshua Bengio. Learning phrase representations using RNN encoder decoder for statistical machine translation. In Proc. of EMNLP, Michael Collins. Log-linear models, MEMMs, and CRFs, URL Stephen Della Pietra, Vincent Della Pietra, and John Lafferty. Inducing features of random fields. IEEE Transactions on Pattern Analysis and Machine Intelligence, 19(4): , Jacob Eisenstein, Amr Ahmed, and Eric P Xing. Sparse additive generative models of text. In Proc. of ICML, Yoav Goldberg. A primer on neural network models for natural language processing, URL Rafal Jozefowicz, Wojciech Zaremba, and Ilya Sutskever. An empirical exploration of recurrent network architectures. In Proc. of ICML, URL Daniel Jurafsky and James H. Martin. N-grams (draft chapter), URL / 102
102 References II Raymond Lau, Ronald Rosenfeld, and Salim Roukos. Trigger-based language models: A maximum entropy approach. In Proc. of ICASSP, Tomas Mikolov, Martin Karafiát, Lukas Burget, Jan Cernockỳ, and Sanjeev Khudanpur. Recurrent neural network based language model. In Proc. of Interspeech, URL http: // Tomas Mikolov, Kai Chen, Greg Corrado, and Jeffrey Dean. Efficient estimation of word representations in vector space. In Proceedings of ICLR, URL Tomas Mikolov, Armand Joulin, Sumit Chopra, Michael Mathieu, and Marc Aurelio Ranzato. Learning longer memory in recurrent neural networks, arxiv: Andriy Mnih and Geoffrey Hinton. Three new graphical models for statistical language modelling. In Proc. of ICML, Noah A. Smith. Probabilistic language models 1.0, URL Martin Sundermeyer, Ralf Schlüter, and Hermann Ney. LSTM neural networks for language modeling. In Proc. of Interspeech, Yulia Tsvetkov, Manaal Faruqui, Wang Ling, Guillaume Lample, and Chris Dyer. Evaluation of word vector representations by subspace alignment. In Proc. of EMNLP, Peter D. Turney and Patrick Pantel. From frequency to meaning: Vector space models of semantics. Journal of Artificial Intelligence Research, 37(1): , URL / 102
Neural Network Language Modeling
Neural Network Language Modeling Instructor: Wei Xu Ohio State University CSE 5525 Many slides from Marek Rei, Philipp Koehn and Noah Smith Course Project Sign up your course project In-class presentation
More informationImproved Learning through Augmenting the Loss
Improved Learning through Augmenting the Loss Hakan Inan inanh@stanford.edu Khashayar Khosravi khosravi@stanford.edu Abstract We present two improvements to the well-known Recurrent Neural Network Language
More informationNatural Language Processing (CSEP 517): Machine Translation (Continued), Summarization, & Finale
Natural Language Processing (CSEP 517): Machine Translation (Continued), Summarization, & Finale Noah Smith c 2017 University of Washington nasmith@cs.washington.edu May 22, 2017 1 / 30 To-Do List Online
More informationDeep learning for Natural Language Processing and Machine Translation
Deep learning for Natural Language Processing and Machine Translation 2015.10.16 Seung-Hoon Na Contents Introduction: Neural network, deep learning Deep learning for Natural language processing Neural
More informationDeep Learning Recurrent Networks 2/28/2018
Deep Learning Recurrent Networks /8/8 Recap: Recurrent networks can be incredibly effective Story so far Y(t+) Stock vector X(t) X(t+) X(t+) X(t+) X(t+) X(t+5) X(t+) X(t+7) Iterated structures are good
More informationCSC321 Lecture 10 Training RNNs
CSC321 Lecture 10 Training RNNs Roger Grosse and Nitish Srivastava February 23, 2015 Roger Grosse and Nitish Srivastava CSC321 Lecture 10 Training RNNs February 23, 2015 1 / 18 Overview Last time, we saw
More informationFrom perceptrons to word embeddings. Simon Šuster University of Groningen
From perceptrons to word embeddings Simon Šuster University of Groningen Outline A basic computational unit Weighting some input to produce an output: classification Perceptron Classify tweets Written
More informationEmpirical Methods in Natural Language Processing Lecture 10a More smoothing and the Noisy Channel Model
Empirical Methods in Natural Language Processing Lecture 10a More smoothing and the Noisy Channel Model (most slides from Sharon Goldwater; some adapted from Philipp Koehn) 5 October 2016 Nathan Schneider
More informationBased on the original slides of Hung-yi Lee
Based on the original slides of Hung-yi Lee New Activation Function Rectified Linear Unit (ReLU) σ z a a = z Reason: 1. Fast to compute 2. Biological reason a = 0 [Xavier Glorot, AISTATS 11] [Andrew L.
More informationHigh Order LSTM/GRU. Wenjie Luo. January 19, 2016
High Order LSTM/GRU Wenjie Luo January 19, 2016 1 Introduction RNN is a powerful model for sequence data but suffers from gradient vanishing and explosion, thus difficult to be trained to capture long
More informationNatural Language Processing (CSE 490U): Language Models
Natural Language Processing (CSE 490U): Language Models Noah Smith c 2017 University of Washington nasmith@cs.washington.edu January 6 9, 2017 1 / 67 Very Quick Review of Probability Event space (e.g.,
More informationCS230: Lecture 8 Word2Vec applications + Recurrent Neural Networks with Attention
CS23: Lecture 8 Word2Vec applications + Recurrent Neural Networks with Attention Today s outline We will learn how to: I. Word Vector Representation i. Training - Generalize results with word vectors -
More informationCSC321 Lecture 15: Recurrent Neural Networks
CSC321 Lecture 15: Recurrent Neural Networks Roger Grosse Roger Grosse CSC321 Lecture 15: Recurrent Neural Networks 1 / 26 Overview Sometimes we re interested in predicting sequences Speech-to-text and
More informationInter-document Contextual Language Model
Inter-document Contextual Language Model Quan Hung Tran and Ingrid Zukerman and Gholamreza Haffari Faculty of Information Technology Monash University, Australia hung.tran,ingrid.zukerman,gholamreza.haffari@monash.edu
More informationword2vec Parameter Learning Explained
word2vec Parameter Learning Explained Xin Rong ronxin@umich.edu Abstract The word2vec model and application by Mikolov et al. have attracted a great amount of attention in recent two years. The vector
More informationLecture 5 Neural models for NLP
CS546: Machine Learning in NLP (Spring 2018) http://courses.engr.illinois.edu/cs546/ Lecture 5 Neural models for NLP Julia Hockenmaier juliahmr@illinois.edu 3324 Siebel Center Office hours: Tue/Thu 2pm-3pm
More informationNatural Language Processing and Recurrent Neural Networks
Natural Language Processing and Recurrent Neural Networks Pranay Tarafdar October 19 th, 2018 Outline Introduction to NLP Word2vec RNN GRU LSTM Demo What is NLP? Natural Language? : Huge amount of information
More informationarxiv: v1 [cs.cl] 21 May 2017
Spelling Correction as a Foreign Language Yingbo Zhou yingbzhou@ebay.com Utkarsh Porwal uporwal@ebay.com Roberto Konow rkonow@ebay.com arxiv:1705.07371v1 [cs.cl] 21 May 2017 Abstract In this paper, we
More informationarxiv: v1 [cs.ne] 19 Dec 2016
A RECURRENT NEURAL NETWORK WITHOUT CHAOS Thomas Laurent Department of Mathematics Loyola Marymount University Los Angeles, CA 90045, USA tlaurent@lmu.edu James von Brecht Department of Mathematics California
More informationa) b) (Natural Language Processing; NLP) (Deep Learning) Bag of words White House RGB [1] IBM
c 1. (Natural Language Processing; NLP) (Deep Learning) RGB IBM 135 8511 5 6 52 yutat@jp.ibm.com a) b) 2. 1 0 2 1 Bag of words White House 2 [1] 2015 4 Copyright c by ORSJ. Unauthorized reproduction of
More informationA QUESTION ANSWERING SYSTEM USING ENCODER-DECODER, SEQUENCE-TO-SEQUENCE, RECURRENT NEURAL NETWORKS. A Project. Presented to
A QUESTION ANSWERING SYSTEM USING ENCODER-DECODER, SEQUENCE-TO-SEQUENCE, RECURRENT NEURAL NETWORKS A Project Presented to The Faculty of the Department of Computer Science San José State University In
More informationNeural Networks with Applications to Vision and Language. Feedforward Networks. Marco Kuhlmann
Neural Networks with Applications to Vision and Language Feedforward Networks Marco Kuhlmann Feedforward networks Linear separability x 2 x 2 0 1 0 1 0 0 x 1 1 0 x 1 linearly separable not linearly separable
More informationEE-559 Deep learning LSTM and GRU
EE-559 Deep learning 11.2. LSTM and GRU François Fleuret https://fleuret.org/ee559/ Mon Feb 18 13:33:24 UTC 2019 ÉCOLE POLYTECHNIQUE FÉDÉRALE DE LAUSANNE The Long-Short Term Memory unit (LSTM) by Hochreiter
More informationFoundations of Natural Language Processing Lecture 5 More smoothing and the Noisy Channel Model
Foundations of Natural Language Processing Lecture 5 More smoothing and the Noisy Channel Model Alex Lascarides (Slides based on those from Alex Lascarides, Sharon Goldwater and Philipop Koehn) 30 January
More informationRandom Coattention Forest for Question Answering
Random Coattention Forest for Question Answering Jheng-Hao Chen Stanford University jhenghao@stanford.edu Ting-Po Lee Stanford University tingpo@stanford.edu Yi-Chun Chen Stanford University yichunc@stanford.edu
More informationNatural Language Processing (CSE 517): Cotext Models (II)
Natural Language Processing (CSE 517): Cotext Models (II) Noah Smith c 2016 University of Washington nasmith@cs.washington.edu January 25, 2016 Thanks to David Mimno for comments. 1 / 30 Three Kinds of
More informationRecurrent Neural Networks (Part - 2) Sumit Chopra Facebook
Recurrent Neural Networks (Part - 2) Sumit Chopra Facebook Recap Standard RNNs Training: Backpropagation Through Time (BPTT) Application to sequence modeling Language modeling Applications: Automatic speech
More informationLecture 6: Neural Networks for Representing Word Meaning
Lecture 6: Neural Networks for Representing Word Meaning Mirella Lapata School of Informatics University of Edinburgh mlap@inf.ed.ac.uk February 7, 2017 1 / 28 Logistic Regression Input is a feature vector,
More informationIntroduction to RNNs!
Introduction to RNNs Arun Mallya Best viewed with Computer Modern fonts installed Outline Why Recurrent Neural Networks (RNNs)? The Vanilla RNN unit The RNN forward pass Backpropagation refresher The RNN
More informationDeep Learning Sequence to Sequence models: Attention Models. 17 March 2018
Deep Learning Sequence to Sequence models: Attention Models 17 March 2018 1 Sequence-to-sequence modelling Problem: E.g. A sequence X 1 X N goes in A different sequence Y 1 Y M comes out Speech recognition:
More informationRecurrent Neural Networks Deep Learning Lecture 5. Efstratios Gavves
Recurrent Neural Networks Deep Learning Lecture 5 Efstratios Gavves Sequential Data So far, all tasks assumed stationary data Neither all data, nor all tasks are stationary though Sequential Data: Text
More informationSemantics with Dense Vectors. Reference: D. Jurafsky and J. Martin, Speech and Language Processing
Semantics with Dense Vectors Reference: D. Jurafsky and J. Martin, Speech and Language Processing 1 Semantics with Dense Vectors We saw how to represent a word as a sparse vector with dimensions corresponding
More informationLecture 17: Neural Networks and Deep Learning
UVA CS 6316 / CS 4501-004 Machine Learning Fall 2016 Lecture 17: Neural Networks and Deep Learning Jack Lanchantin Dr. Yanjun Qi 1 Neurons 1-Layer Neural Network Multi-layer Neural Network Loss Functions
More informationSequence Modeling with Neural Networks
Sequence Modeling with Neural Networks Harini Suresh y 0 y 1 y 2 s 0 s 1 s 2... x 0 x 1 x 2 hat is a sequence? This morning I took the dog for a walk. sentence medical signals speech waveform Successes
More informationNEURAL LANGUAGE MODELS
COMP90042 LECTURE 14 NEURAL LANGUAGE MODELS LANGUAGE MODELS Assign a probability to a sequence of words Framed as sliding a window over the sentence, predicting each word from finite context to left E.g.,
More informationNatural Language Processing (CSEP 517): Text Classification
Natural Language Processing (CSEP 517): Text Classification Noah Smith c 2017 University of Washington nasmith@cs.washington.edu April 10, 2017 1 / 71 To-Do List Online quiz: due Sunday Read: Jurafsky
More informationNeural Networks Language Models
Neural Networks Language Models Philipp Koehn 10 October 2017 N-Gram Backoff Language Model 1 Previously, we approximated... by applying the chain rule p(w ) = p(w 1, w 2,..., w n ) p(w ) = i p(w i w 1,...,
More informationNatural Language Processing (CSE 517): Cotext Models
Natural Language Processing (CSE 517): Cotext Models Noah Smith c 2018 University of Washington nasmith@cs.washington.edu April 18, 2018 1 / 32 Latent Dirichlet Allocation (Blei et al., 2003) Widely used
More informationFaster Training of Very Deep Networks Via p-norm Gates
Faster Training of Very Deep Networks Via p-norm Gates Trang Pham, Truyen Tran, Dinh Phung, Svetha Venkatesh Center for Pattern Recognition and Data Analytics Deakin University, Geelong Australia Email:
More informationNatural Language Processing
Natural Language Processing Info 159/259 Lecture 7: Language models 2 (Sept 14, 2017) David Bamman, UC Berkeley Language Model Vocabulary V is a finite set of discrete symbols (e.g., words, characters);
More informationRECURRENT NEURAL NETWORK BASED LANGUAGE MODELING WITH CONTROLLABLE EXTERNAL MEMORY
RECURRENT NEURAL NETWORK BASED LANGUAGE MODELING WITH CONTROLLABLE EXTERNAL MEMORY Wei-Jen Ko 1, Bo-Hsiang Tseng 2, Hung-Yi Lee 2 1 Research Center for Information Technology Innovation, Academia Sinica
More informationarxiv: v3 [cs.cl] 30 Jan 2016
word2vec Parameter Learning Explained Xin Rong ronxin@umich.edu arxiv:1411.2738v3 [cs.cl] 30 Jan 2016 Abstract The word2vec model and application by Mikolov et al. have attracted a great amount of attention
More informationIMPROVING STOCHASTIC GRADIENT DESCENT
IMPROVING STOCHASTIC GRADIENT DESCENT WITH FEEDBACK Jayanth Koushik & Hiroaki Hayashi Language Technologies Institute Carnegie Mellon University Pittsburgh, PA 15213, USA {jkoushik,hiroakih}@cs.cmu.edu
More informationDISTRIBUTIONAL SEMANTICS
COMP90042 LECTURE 4 DISTRIBUTIONAL SEMANTICS LEXICAL DATABASES - PROBLEMS Manually constructed Expensive Human annotation can be biased and noisy Language is dynamic New words: slangs, terminology, etc.
More informationMachine Translation. 10: Advanced Neural Machine Translation Architectures. Rico Sennrich. University of Edinburgh. R. Sennrich MT / 26
Machine Translation 10: Advanced Neural Machine Translation Architectures Rico Sennrich University of Edinburgh R. Sennrich MT 2018 10 1 / 26 Today s Lecture so far today we discussed RNNs as encoder and
More informationGated Recurrent Neural Tensor Network
Gated Recurrent Neural Tensor Network Andros Tjandra, Sakriani Sakti, Ruli Manurung, Mirna Adriani and Satoshi Nakamura Faculty of Computer Science, Universitas Indonesia, Indonesia Email: andros.tjandra@gmail.com,
More informationarxiv: v1 [stat.ml] 18 Nov 2017
MinimalRNN: Toward More Interpretable and Trainable Recurrent Neural Networks arxiv:1711.06788v1 [stat.ml] 18 Nov 2017 Minmin Chen Google Mountain view, CA 94043 minminc@google.com Abstract We introduce
More informationSparse vectors recap. ANLP Lecture 22 Lexical Semantics with Dense Vectors. Before density, another approach to normalisation.
ANLP Lecture 22 Lexical Semantics with Dense Vectors Henry S. Thompson Based on slides by Jurafsky & Martin, some via Dorota Glowacka 5 November 2018 Previous lectures: Sparse vectors recap How to represent
More informationANLP Lecture 22 Lexical Semantics with Dense Vectors
ANLP Lecture 22 Lexical Semantics with Dense Vectors Henry S. Thompson Based on slides by Jurafsky & Martin, some via Dorota Glowacka 5 November 2018 Henry S. Thompson ANLP Lecture 22 5 November 2018 Previous
More informationRecurrent Neural Networks 2. CS 287 (Based on Yoav Goldberg s notes)
Recurrent Neural Networks 2 CS 287 (Based on Yoav Goldberg s notes) Review: Representation of Sequence Many tasks in NLP involve sequences w 1,..., w n Representations as matrix dense vectors X (Following
More informationDeep Learning for Natural Language Processing. Sidharth Mudgal April 4, 2017
Deep Learning for Natural Language Processing Sidharth Mudgal April 4, 2017 Table of contents 1. Intro 2. Word Vectors 3. Word2Vec 4. Char Level Word Embeddings 5. Application: Entity Matching 6. Conclusion
More informationIntroduction to Deep Neural Networks
Introduction to Deep Neural Networks Presenter: Chunyuan Li Pattern Classification and Recognition (ECE 681.01) Duke University April, 2016 Outline 1 Background and Preliminaries Why DNNs? Model: Logistic
More informationCombining Static and Dynamic Information for Clinical Event Prediction
Combining Static and Dynamic Information for Clinical Event Prediction Cristóbal Esteban 1, Antonio Artés 2, Yinchong Yang 1, Oliver Staeck 3, Enrique Baca-García 4 and Volker Tresp 1 1 Siemens AG and
More informationRecurrent Neural Networks. Jian Tang
Recurrent Neural Networks Jian Tang tangjianpku@gmail.com 1 RNN: Recurrent neural networks Neural networks for sequence modeling Summarize a sequence with fix-sized vector through recursively updating
More informationConditional Language Modeling. Chris Dyer
Conditional Language Modeling Chris Dyer Unconditional LMs A language model assigns probabilities to sequences of words,. w =(w 1,w 2,...,w`) It is convenient to decompose this probability using the chain
More informationMachine Learning. Neural Networks. (slides from Domingos, Pardo, others)
Machine Learning Neural Networks (slides from Domingos, Pardo, others) For this week, Reading Chapter 4: Neural Networks (Mitchell, 1997) See Canvas For subsequent weeks: Scaling Learning Algorithms toward
More informationAn overview of word2vec
An overview of word2vec Benjamin Wilson Berlin ML Meetup, July 8 2014 Benjamin Wilson word2vec Berlin ML Meetup 1 / 25 Outline 1 Introduction 2 Background & Significance 3 Architecture 4 CBOW word representations
More informationNeed for Deep Networks Perceptron. Can only model linear functions. Kernel Machines. Non-linearity provided by kernels
Need for Deep Networks Perceptron Can only model linear functions Kernel Machines Non-linearity provided by kernels Need to design appropriate kernels (possibly selecting from a set, i.e. kernel learning)
More informationarxiv: v2 [cs.cl] 1 Jan 2019
Variational Self-attention Model for Sentence Representation arxiv:1812.11559v2 [cs.cl] 1 Jan 2019 Qiang Zhang 1, Shangsong Liang 2, Emine Yilmaz 1 1 University College London, London, United Kingdom 2
More informationStructured Neural Networks (I)
Structured Neural Networks (I) CS 690N, Spring 208 Advanced Natural Language Processing http://peoplecsumassedu/~brenocon/anlp208/ Brendan O Connor College of Information and Computer Sciences University
More informationPart-of-Speech Tagging + Neural Networks 3: Word Embeddings CS 287
Part-of-Speech Tagging + Neural Networks 3: Word Embeddings CS 287 Review: Neural Networks One-layer multi-layer perceptron architecture, NN MLP1 (x) = g(xw 1 + b 1 )W 2 + b 2 xw + b; perceptron x is the
More informationAlgorithms for NLP. Language Modeling III. Taylor Berg-Kirkpatrick CMU Slides: Dan Klein UC Berkeley
Algorithms for NLP Language Modeling III Taylor Berg-Kirkpatrick CMU Slides: Dan Klein UC Berkeley Announcements Office hours on website but no OH for Taylor until next week. Efficient Hashing Closed address
More informationArtificial Neural Networks D B M G. Data Base and Data Mining Group of Politecnico di Torino. Elena Baralis. Politecnico di Torino
Artificial Neural Networks Data Base and Data Mining Group of Politecnico di Torino Elena Baralis Politecnico di Torino Artificial Neural Networks Inspired to the structure of the human brain Neurons as
More informationDeep Gate Recurrent Neural Network
JMLR: Workshop and Conference Proceedings 63:350 365, 2016 ACML 2016 Deep Gate Recurrent Neural Network Yuan Gao University of Helsinki Dorota Glowacka University of Helsinki gaoyuankidult@gmail.com glowacka@cs.helsinki.fi
More informationRecurrent Neural Networks. COMP-550 Oct 5, 2017
Recurrent Neural Networks COMP-550 Oct 5, 2017 Outline Introduction to neural networks and deep learning Feedforward neural networks Recurrent neural networks 2 Classification Review y = f( x) output label
More informationMachine Learning. Neural Networks. (slides from Domingos, Pardo, others)
Machine Learning Neural Networks (slides from Domingos, Pardo, others) Human Brain Neurons Input-Output Transformation Input Spikes Output Spike Spike (= a brief pulse) (Excitatory Post-Synaptic Potential)
More informationCSC321 Lecture 15: Exploding and Vanishing Gradients
CSC321 Lecture 15: Exploding and Vanishing Gradients Roger Grosse Roger Grosse CSC321 Lecture 15: Exploding and Vanishing Gradients 1 / 23 Overview Yesterday, we saw how to compute the gradient descent
More informationRecurrent Neural Networks (RNN) and Long-Short-Term-Memory (LSTM) Yuan YAO HKUST
1 Recurrent Neural Networks (RNN) and Long-Short-Term-Memory (LSTM) Yuan YAO HKUST Summary We have shown: Now First order optimization methods: GD (BP), SGD, Nesterov, Adagrad, ADAM, RMSPROP, etc. Second
More informationRecurrent Neural Networks with Flexible Gates using Kernel Activation Functions
2018 IEEE International Workshop on Machine Learning for Signal Processing (MLSP 18) Recurrent Neural Networks with Flexible Gates using Kernel Activation Functions Authors: S. Scardapane, S. Van Vaerenbergh,
More informationGraphical models for part of speech tagging
Indian Institute of Technology, Bombay and Research Division, India Research Lab Graphical models for part of speech tagging Different Models for POS tagging HMM Maximum Entropy Markov Models Conditional
More informationEve: A Gradient Based Optimization Method with Locally and Globally Adaptive Learning Rates
Eve: A Gradient Based Optimization Method with Locally and Globally Adaptive Learning Rates Hiroaki Hayashi 1,* Jayanth Koushik 1,* Graham Neubig 1 arxiv:1611.01505v3 [cs.lg] 11 Jun 2018 Abstract Adaptive
More informationarxiv: v1 [cs.cl] 31 May 2015
Recurrent Neural Networks with External Memory for Language Understanding Baolin Peng 1, Kaisheng Yao 2 1 The Chinese University of Hong Kong 2 Microsoft Research blpeng@se.cuhk.edu.hk, kaisheny@microsoft.com
More informationLecture 15: Recurrent Neural Nets
Lecture 15: Recurrent Neural Nets Roger Grosse 1 Introduction Most of the prediction tasks we ve looked at have involved pretty simple kinds of outputs, such as real values or discrete categories. But
More informationMachine Learning for Large-Scale Data Analysis and Decision Making A. Neural Networks Week #6
Machine Learning for Large-Scale Data Analysis and Decision Making 80-629-17A Neural Networks Week #6 Today Neural Networks A. Modeling B. Fitting C. Deep neural networks Today s material is (adapted)
More informationFeedforward Neural Networks
Feedforward Neural Networks Michael Collins 1 Introduction In the previous notes, we introduced an important class of models, log-linear models. In this note, we describe feedforward neural networks, which
More informationRecurrent and Recursive Networks
Neural Networks with Applications to Vision and Language Recurrent and Recursive Networks Marco Kuhlmann Introduction Applications of sequence modelling Map unsegmented connected handwriting to strings.
More informationDeep Learning for NLP
Deep Learning for NLP CS224N Christopher Manning (Many slides borrowed from ACL 2012/NAACL 2013 Tutorials by me, Richard Socher and Yoshua Bengio) Machine Learning and NLP NER WordNet Usually machine learning
More informationCharacter-level Language Modeling with Gated Hierarchical Recurrent Neural Networks
Interspeech 2018 2-6 September 2018, Hyderabad Character-level Language Modeling with Gated Hierarchical Recurrent Neural Networks Iksoo Choi, Jinhwan Park, Wonyong Sung Department of Electrical and Computer
More informationLong-Short Term Memory and Other Gated RNNs
Long-Short Term Memory and Other Gated RNNs Sargur Srihari srihari@buffalo.edu This is part of lecture slides on Deep Learning: http://www.cedar.buffalo.edu/~srihari/cse676 1 Topics in Sequence Modeling
More informationNeural Turing Machine. Author: Alex Graves, Greg Wayne, Ivo Danihelka Presented By: Tinghui Wang (Steve)
Neural Turing Machine Author: Alex Graves, Greg Wayne, Ivo Danihelka Presented By: Tinghui Wang (Steve) Introduction Neural Turning Machine: Couple a Neural Network with external memory resources The combined
More informationECE521 Lecture 7/8. Logistic Regression
ECE521 Lecture 7/8 Logistic Regression Outline Logistic regression (Continue) A single neuron Learning neural networks Multi-class classification 2 Logistic regression The output of a logistic regression
More information11/3/15. Deep Learning for NLP. Deep Learning and its Architectures. What is Deep Learning? Advantages of Deep Learning (Part 1)
11/3/15 Machine Learning and NLP Deep Learning for NLP Usually machine learning works well because of human-designed representations and input features CS224N WordNet SRL Parser Machine learning becomes
More informationDeep Learning for NLP Part 2
Deep Learning for NLP Part 2 CS224N Christopher Manning (Many slides borrowed from ACL 2012/NAACL 2013 Tutorials by me, Richard Socher and Yoshua Bengio) 2 Part 1.3: The Basics Word Representations The
More informationDeep Feedforward Networks. Seung-Hoon Na Chonbuk National University
Deep Feedforward Networks Seung-Hoon Na Chonbuk National University Neural Network: Types Feedforward neural networks (FNN) = Deep feedforward networks = multilayer perceptrons (MLP) No feedback connections
More informationDeep Learning. Recurrent Neural Network (RNNs) Ali Ghodsi. October 23, Slides are partially based on Book in preparation, Deep Learning
Recurrent Neural Network (RNNs) University of Waterloo October 23, 2015 Slides are partially based on Book in preparation, by Bengio, Goodfellow, and Aaron Courville, 2015 Sequential data Recurrent neural
More informationLong Short- Term Memory (LSTM) M1 Yuichiro Sawai Computa;onal Linguis;cs Lab. January 15, Deep Lunch
Long Short- Term Memory (LSTM) M1 Yuichiro Sawai Computa;onal Linguis;cs Lab. January 15, 2015 @ Deep Lunch 1 Why LSTM? OJen used in many recent RNN- based systems Machine transla;on Program execu;on Can
More informationDeep Learning Basics Lecture 8: Autoencoder & DBM. Princeton University COS 495 Instructor: Yingyu Liang
Deep Learning Basics Lecture 8: Autoencoder & DBM Princeton University COS 495 Instructor: Yingyu Liang Autoencoder Autoencoder Neural networks trained to attempt to copy its input to its output Contain
More informationDEEP LEARNING AND NEURAL NETWORKS: BACKGROUND AND HISTORY
DEEP LEARNING AND NEURAL NETWORKS: BACKGROUND AND HISTORY 1 On-line Resources http://neuralnetworksanddeeplearning.com/index.html Online book by Michael Nielsen http://matlabtricks.com/post-5/3x3-convolution-kernelswith-online-demo
More informationNeed for Deep Networks Perceptron. Can only model linear functions. Kernel Machines. Non-linearity provided by kernels
Need for Deep Networks Perceptron Can only model linear functions Kernel Machines Non-linearity provided by kernels Need to design appropriate kernels (possibly selecting from a set, i.e. kernel learning)
More informationLecture 15: Exploding and Vanishing Gradients
Lecture 15: Exploding and Vanishing Gradients Roger Grosse 1 Introduction Last lecture, we introduced RNNs and saw how to derive the gradients using backprop through time. In principle, this lets us train
More informationBetter Conditional Language Modeling. Chris Dyer
Better Conditional Language Modeling Chris Dyer Conditional LMs A conditional language model assigns probabilities to sequences of words, w =(w 1,w 2,...,w`), given some conditioning context, x. As with
More informationLogistic Regression & Neural Networks
Logistic Regression & Neural Networks CMSC 723 / LING 723 / INST 725 Marine Carpuat Slides credit: Graham Neubig, Jacob Eisenstein Logistic Regression Perceptron & Probabilities What if we want a probability
More informationStephen Scott.
1 / 35 (Adapted from Vinod Variyam and Ian Goodfellow) sscott@cse.unl.edu 2 / 35 All our architectures so far work on fixed-sized inputs neural networks work on sequences of inputs E.g., text, biological
More informationDeep Learning. Ali Ghodsi. University of Waterloo
University of Waterloo Language Models A language model computes a probability for a sequence of words: P(w 1,..., w T ) Useful for machine translation Word ordering: p (the cat is small) > p (small the
More informationLanguage Modeling. Hung-yi Lee 李宏毅
Language Modeling Hung-yi Lee 李宏毅 Language modeling Language model: Estimated the probability of word sequence Word sequence: w 1, w 2, w 3,., w n P(w 1, w 2, w 3,., w n ) Application: speech recognition
More informationTTIC 31230, Fundamentals of Deep Learning David McAllester, April Sequence to Sequence Models and Attention
TTIC 31230, Fundamentals of Deep Learning David McAllester, April 2017 Sequence to Sequence Models and Attention Encode-Decode Architectures for Machine Translation [Figure from Luong et al.] In Sutskever
More informationACCELERATING RECURRENT NEURAL NETWORK TRAINING VIA TWO STAGE CLASSES AND PARALLELIZATION
ACCELERATING RECURRENT NEURAL NETWORK TRAINING VIA TWO STAGE CLASSES AND PARALLELIZATION Zhiheng Huang, Geoffrey Zweig, Michael Levit, Benoit Dumoulin, Barlas Oguz and Shawn Chang Speech at Microsoft,
More informationDeep learning / Ian Goodfellow, Yoshua Bengio and Aaron Courville. - Cambridge, MA ; London, Spis treści
Deep learning / Ian Goodfellow, Yoshua Bengio and Aaron Courville. - Cambridge, MA ; London, 2017 Spis treści Website Acknowledgments Notation xiii xv xix 1 Introduction 1 1.1 Who Should Read This Book?
More informationRecurrent neural networks
12-1: Recurrent neural networks Prof. J.C. Kao, UCLA Recurrent neural networks Motivation Network unrollwing Backpropagation through time Vanishing and exploding gradients LSTMs GRUs 12-2: Recurrent neural
More informationLecture 13: Discriminative Sequence Models (MEMM and Struct. Perceptron)
Lecture 13: Discriminative Sequence Models (MEMM and Struct. Perceptron) Intro to NLP, CS585, Fall 2014 http://people.cs.umass.edu/~brenocon/inlp2014/ Brendan O Connor (http://brenocon.com) 1 Models for
More information