Weight Moving Average = n

Similar documents
First come, first served (FCFS) Batch

Optimization Methods: Linear Programming Applications Assignment Problem 1. Module 4 Lecture Notes 3. Assignment Problem

A NEW APPROACH TO SOLVE AN UNBALANCED ASSIGNMENT PROBLEM

Optimization Methods MIT 2.098/6.255/ Final exam

Integer Programming (IP)

Definitions and Theorems. where x are the decision variables. c, b, and a are constant coefficients.

Linear Programming and the Simplex Method

Linear Programming! References! Introduction to Algorithms.! Dasgupta, Papadimitriou, Vazirani. Algorithms.! Cormen, Leiserson, Rivest, and Stein.

Queuing Theory. Basic properties, Markovian models, Networks of queues, General service time distributions, Finite source models, Multiserver queues

Integer Linear Programming

TRANSPORTATION AND ASSIGNMENT PROBLEMS

Markov Decision Processes

w (1) ˆx w (1) x (1) /ρ and w (2) ˆx w (2) x (2) /ρ.

There is no straightforward approach for choosing the warmup period l.

SINGLE-CHANNEL QUEUING PROBLEMS APPROACH

B. Maddah ENMG 622 ENMG /27/07

Axioms of Measure Theory

The Simplex algorithm: Introductory example. The Simplex algorithm: Introductory example (2)

c. Explain the basic Newsvendor model. Why is it useful for SC models? e. What additional research do you believe will be helpful in this area?

Linear Regression Demystified

Optimally Sparse SVMs

Scheduling under Uncertainty using MILP Sensitivity Analysis

Chapter 1 Simple Linear Regression (part 6: matrix version)

Chapter Vectors

U8L1: Sec Equations of Lines in R 2

ACO Comprehensive Exam 9 October 2007 Student code A. 1. Graph Theory

Announcements. Queueing Systems: Lecture 1. Lecture Outline. Topics in Queueing Theory

Chapter 15 MSM2M2 Introduction To Management Mathematics

Statistical Fundamentals and Control Charts

CS/ECE 715 Spring 2004 Homework 5 (Due date: March 16)

CS284A: Representations and Algorithms in Molecular Biology

Sequences, Mathematical Induction, and Recursion. CSE 2353 Discrete Computational Structures Spring 2018

( ) = p and P( i = b) = q.

Kinetics of Complex Reactions

LINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity

Properties and Hypothesis Testing

UNIVERSITY OF NORTHERN COLORADO MATHEMATICS CONTEST. First Round For all Colorado Students Grades 7-12 November 3, 2007

Mathematical Notation Math Finite Mathematics

Linear regression. Daniel Hsu (COMS 4771) (y i x T i β)2 2πσ. 2 2σ 2. 1 n. (x T i β y i ) 2. 1 ˆβ arg min. β R n d

Lecture 24: Variable selection in linear models

MATHEMATICAL SCIENCES PAPER-II

Chapter 6 Sampling Distributions

LINEAR PROGRAMMING II

Mathematical Statistics - MS

The Random Walk For Dummies

Estimation for Complete Data

SECTION 1.5 : SUMMATION NOTATION + WORK WITH SEQUENCES

Inverse Matrix. A meaning that matrix B is an inverse of matrix A.

1. By using truth tables prove that, for all statements P and Q, the statement

A queueing system can be described as customers arriving for service, waiting for service if it is not immediate, and if having waited for service,

Lecture 5: April 17, 2013

Chapter 4 - Summarizing Numerical Data

Introduction to Optimization Techniques. How to Solve Equations

Economics 250 Assignment 1 Suggested Answers. 1. We have the following data set on the lengths (in minutes) of a sample of long-distance phone calls

MATHEMATICAL SCIENCES

MASSACHUSETTS INSTITUTE OF TECHNOLOGY

(A sequence also can be thought of as the list of function values attained for a function f :ℵ X, where f (n) = x n for n 1.) x 1 x N +k x N +4 x 3

Algebra of Least Squares

Randomized Algorithms I, Spring 2018, Department of Computer Science, University of Helsinki Homework 1: Solutions (Discussed January 25, 2018)

t distribution [34] : used to test a mean against an hypothesized value (H 0 : µ = µ 0 ) or the difference

SEQUENCE AND SERIES NCERT

Differentiable Convex Functions

September 2012 C1 Note. C1 Notes (Edexcel) Copyright - For AS, A2 notes and IGCSE / GCSE worksheets 1

MATH 320: Probability and Statistics 9. Estimation and Testing of Parameters. Readings: Pruim, Chapter 4

Queueing theory and Replacement model

Applications in Linear Algebra and Uses of Technology

Review Problems 1. ICME and MS&E Refresher Course September 19, 2011 B = C = AB = A = A 2 = A 3... C 2 = C 3 = =

Essential Question How can you recognize an arithmetic sequence from its graph?

ECON 3150/4150, Spring term Lecture 3

Advanced Algebra SS Semester 2 Final Exam Study Guide Mrs. Dunphy

Massachusetts Institute of Technology

10-701/ Machine Learning Mid-term Exam Solution

Product measures, Tonelli s and Fubini s theorems For use in MAT3400/4400, autumn 2014 Nadia S. Larsen. Version of 13 October 2014.

CSE 202 Homework 1 Matthias Springer, A Yes, there does always exist a perfect matching without a strong instability.

subject to A 1 x + A 2 y b x j 0, j = 1,,n 1 y j = 0 or 1, j = 1,,n 2

2.4 - Sequences and Series

First, note that the LS residuals are orthogonal to the regressors. X Xb X y = 0 ( normal equations ; (k 1) ) So,

CLRM estimation Pietro Coretto Econometrics

Algorithms for Clustering

Simulation of Discrete Event Systems

Regression, Inference, and Model Building

Increasing timing capacity using packet coloring

The Method of Least Squares. To understand least squares fitting of data.

CS 171 Lecture Outline October 09, 2008

SNAP Centre Workshop. Basic Algebraic Manipulation

Math 475, Problem Set #12: Answers

Complex Numbers Solutions

Chapter 11 Output Analysis for a Single Model. Banks, Carson, Nelson & Nicol Discrete-Event System Simulation

CEE 522 Autumn Uncertainty Concepts for Geotechnical Engineering

Control chart for number of customers in the system of M [X] / M / 1 Queueing system

AH Checklist (Unit 3) AH Checklist (Unit 3) Matrices

Lecture 6 Chi Square Distribution (χ 2 ) and Least Squares Fitting

Chapter 6. Advanced Counting Techniques

Machine Learning Brett Bernstein

4. Basic probability theory

The Boolean Ring of Intervals

6.867 Machine learning, lecture 7 (Jaakkola) 1

USA Mathematical Talent Search Round 3 Solutions Year 27 Academic Year

Recurrence Relations

Statistical Inference (Chapter 10) Statistical inference = learn about a population based on the information provided by a sample.

Transcription:

1

Forecastig Simple Movig Average- F = 1 N (D + D 1 + D 2 + D 3 + ) Movig Weight Average- Weight Movig Average = W id i i=1 W i Sigle (Simple) Expoetial Smoothig- F t = F t 1 + α(d t 1 F t 1 ) or F t = (1 α)f t 1 + αd t 1 if previous forecastig is ot give F t = αd t + α(1 α)d t 1 + α(1 α) 2 D t 2. Where F t = Smoothed average forecast for period t F t 1 =Previous period forecast α=smoothig costat Liear Regressio- Y = a + bx Forecastig Error- Bias- Mea Absolute Deviatio- Mea Square Error- Mea Absolute Percetage Error- y = a + b x xy = x + b x 2 e t = (D t F t ) Bias = 1 N (D t F t ) MAD = 1 N D t F t MSE = 1 N (D t F t ) 2 MAPE = 1 N D t F t 100 Ivetory If D= demad/year, C o =Order cost, C c = Carryig cost, P= Purchase price/uit Q =Ecoomic Order Quatity, K= Productio Rate ad C S =Shortage Cost/uit/period Case-1 Purchase Model With Istataeous Repleishmet ad Without Shortage- 1. EOQ Q = 2DC o C c 2. No. of order = D Q D t at EOQ Ivertory cost = Order cost 3. Time Take Per Order = Q D 4. Total Cost = Uit cost + Ivertory cost + Order cost = (D P) + ( Q 2 C c) + ( D Q C o)= (D P) + 2DC c C o Case-2 Maufacturig Model Without Shortage- 2

1. EOQ Q = 2C od 2. t 1 = Q K 3. t 2 = Q (1 D K ) D C c (1 D K ) 4. Total optimum cost = 2DC c C o (1 D K ) Case 3 Purchase Model With Shortage- 1. Q = EOQ = 2DC o ( C s+c c ) C c C s 2. Q 1 = 2DC o ( C s ), Q C c C s +C 2 = Q Q 1 c 3. t = Q, t D 1 = Q 1 ad t D 2 = Q 2 D 4. Total optimum cost = 2DC c C o ( C s C s +C c ) Case 4 Maufacturig Model With Shortfall 1. Q = EOQ = 2DC o C c (1 D K ) (C s+c c C s ) 2. Q 1 = 2DC o ( C s ) (1 D ) C c C s +C c K 3

3. Q 1 = (1 D K ) Q Q 2 4. t = Q, t D 1 = Q 1 ad t K D 2 = Q 1 D 1. t 3 = Q 3 D ad t 4 = Q 2 K D Lead Time Demad + Safety Stock = Reorder Poit PERT ad CPM EFT = EST + activity time LFT = LST + Duratio of activity Total Float- Free Float- FFo= (Ej-Ei)-Tij Idepedet Float - Example- 1. Total float = L2 (E1 + t12) = 57 (20 + 19) = 18 2. Free float = E2 E1 t12 = 0 3. Idepedet float = E2 (L1 + t12) = -18 PERT Expected time- t e = t 0+4t m +t p 6 1. t0 = Optimistic time i.e., shortest possible time to complete the activity if all goes well. Stadard deviatio- 2. tp = Pessimistic time i.e., logest time that a activity could take if everythig goes wrog. 3. tm = Most likely time i.e., ormal time of a activity would take. Variace - Crashig- Stadard Normal Variatio (SNV)- 4

Liear Programmig Simplex Method Case 1. Maximizatio Problem Max Z = 3x1 + 5x2 s / t 3x1 + 2x2 18 ( I) x1 4 ( II) x2 6 ( III ) x1, x2 0 Stadard Form: Max Z = 3x1 + 5x2 + 0w1 + 0w2 + 0w3 3x1 + 2x2 + w1 + 0w2 + 0w3 = 18 x1 + 0x2 + 0w1 + w2 + 0w3 = 4 0x1 + x2 + 0w1 + 0w2 + w3 = 6 To prepare iitial Table: Table - I Ij = (Zj-cj) = ( aij.ci)-cj Iterpretatio of Simplex Table Table - I Key Colum Mi Ij [ Most Negative ] Key Row Mi positive ratio. How to get ext table? Leavig variable : w3 Eterig variable : x2 Key o. = 1 For old key row : New No.= Old No./key No. For other rows: ( Correspodig Key Row No.). ( Correspodig Key Colum No.) New No. = Old No. Key No. 18 18 - (6*2)/1 = 6 I(w3) = 0 0 - [1*(-5)]/1 = 5 5

Table - II Key Colum Mi Ij Key Row Mi positive ratio Table - III This is the fial Table The Optimal Solutio is x1 = 2, x2 = 6 givig Z = 36 Type of Solutios : Basic, Feasible/Ifeasible, Optimal/ No-Optimal, Uique/Alterative Optimal, Bouded/Ubouded, Degeerate/No-Degeerate Aalysis of Solutio 1. This is a Basic solutio, as values of basic variables are Positive 2. This is a feasible solutio, as values of basic variables, ot cotaiig Artificial Variable, are Positive ad all costraits are satisfied 3. This Feasible solutio is a Optimal, as all values i Idex Row are positive. 4. If there is a Artificial Variable, as Basic variable i fial table, it is called as Ifeasible solutio 5. This solutio is uique Optimal, as the umber of zeroes are equal to umber of basic variables i Idex Row i fial Table. 6. If the umber of zeroes are more tha umber of basic variables i Idex Row i fial Table, it is a case of more tha oe optimal solutios. 7. This is a Bouded Solutio, as the values of all Basic variables i fial table, are fiite positive. 8. This is a No-degeerate Solutio, as value of oe of the basic variables is Zero, i fial table. 9. If value of at least oe of the basic variables is Zero i Idex Row i fial Table, it is a Degeerate Solutio. Duality With Example 1. Case-1 Max.Z = x1 - x2 + 3x3 s/t x1 + x2 + x3 10 2 x1 - x2 - x3 2 2x1-2x2-3x3 6 x1, x2, x3 0 Dual of this would be Mi Z = 10y1 +2y2 + 6y3 s/t y1 + 2 y2 + 2y3 1 y1 - y2-2y3-1 y1 - y2-3y3 3 y1, y3, y3 0, 6

2. Case -2 Mi Z = 20 x1 + 23 x2 s/t - 4 x1 - x2-8 5 x1-3 x2-4 x1, x2 0 Dual of this would be Max Z = 8 y1 + 4 y2 S/t 4 y1-5 y2 20 y1 + 3 y2 23 y1, y2 0, 3. Case-3 Max.Zp = x1+2x2-3x3 s/t 2 x1 + x2 +x3 10 3 x1 - x2 + 2 x3 110(this eed to be coverted i less tha form) x1 + 2x2 - x3 = 4 x1, x3 0, x2 urestricted Dual of this would be Mi. Zd = 10 y1-110 y2 +4 y3 s/t 2 y1-3 y2 + y3 1 y1 + y2 +2 y3 = 2 y1-2 y2 - y3-3 y1, y2 0, y3 urestricted Assigmet Steps of Solvig Assigmet Problem- 1. Idetify the miimum elemet i each row ad subtract it from every elemet of that row. 2. Idetify the miimum elemet i each colum ad subtract it from every elemet of that colum 3. Make the assigmets for the reduced matrix obtaied from steps 1 ad 2 i the followig way: I. For each row or colum with a sigle zero value cell that has ot be assiged or elimiated, box that zero value as a assiged cell. II. For every zero that becomes assiged, cross out (X) all other zeros i the same row ad the same colum. III. If for a row ad a colum, there are two or more zeros ad oe caot be chose by ispectio, choose the cell arbitrarily for assigmet. IV. The above process may be cotiued util every zero cell is either assiged or crossed (X) 4. A optimal assigmet is foud, if the umber of assiged cells equals the umber of rows (ad colums). I case you have chose a zero cell arbitrarily, there may be alterate optimal solutios. If o optimal solutio is foud, go to step 5. 5. Draw the miimum umber of vertical ad horizotal lies ecessary to cover all the zeros i the reduced matrix obtaied from step 3 by adoptig the followig procedure: I. Mark all the rows that do ot have assigmets. II. Mark all the colums (ot already marked) which have zeros i the marked rows. III. Mark all the rows (ot already marked) that have assigmets i marked colums. IV. Repeat steps 5 (ii) ad (iii) util o more rows or colums ca be marked. V. Draw straight lies through all umarked rows ad marked colums. 7

1. Select the smallest elemet (i.e., 1) from all the ucovered elemets. Subtract this smallest elemet from all the ucovered elemets ad add it to the elemets, which lie at the itersectio of two lies. Thus, we obtai aother reduced matrix for fresh assigmet. Note- For maximizatio problem the matrix is coverted ito miimizatio problem by subtractig each elemet by the elemet havig maximum value i the matrix. Queuig Theory Arrival rate is λ ad mea service rate is deoted by µ. Kedall s otatio:- (a/b/c) : (d/e) a = Probability law for the arrival time b = Probability law accordig to which the customers are beig served. c = umber of chaels d = capacity of the system e = queue disciplie. Traffic Itesity( ρ) = Mea arrival rate/ Mea service rate = λ / µ Formulas of Queuig Theory 1. Expected umber of customers i the system (Ls) = λ / (µ - λ) 2. Expected umber of customers i the queue (Lq) = λ 2 / µ (µ - λ) 3. Expected waitig time for a customer i the queue(wq) = λ / µ (µ - λ) 4. Expected waitig time for a customer i the system (Ws) = 1 / (µ - λ) 5. Probability that the queue is o-empty P(>1) =( λ / µ) 2 6. Probability that the umber of customers, i the system exceeds a give umber, P(>=k) = =( λ / µ)k 7. Expected legth of o-empty queue = µ / (µ - λ). Sequecig ad Schedulig Sequecig 1. Johso s Problem 1. 2 machies ad jobs Step 1: Fid the miimum amog various ti1 ad ti2. Step 2a : If the miimum processig time requires machie 1, place the associated job i the first available positio i sequece. Go to Step 3. Step 2b : If the miimum processig time requires machie 2, place the associated job i the last available positio i sequece. Go to Step 3. Step 3: Remove the assiged job from cosideratio ad retur to Step 1 util all positios i sequece are filled. 2. 3 machies ad jobs two coditios of this approach The smallest processig time o machie A is greater tha or equal to the greatest processig time o machie B, i.e., Mi. (Ai) Max. (Bi) The smallest processig time o machie C is greater tha or equal to the greatest processig time o machie B, i.e., Max. (Bi) Mi. (Ci) If either or both of the above coditios are satisfied, the we replace the three machies by two fictitious machies G & H with correspodig processig times give by Gi = Ai + Bi Hi = Bi + Ci Where Gi & Hi are the processig times for ith job o machie G ad H respectively Time Study- Normal time = Observed time x Ratig factor Stadard time = Normal time + allowaces 8

9