Course Completion Tracker
|
|
Course Promo
Management Science is a discipline that attempts to aid managerial decision making by applying a scientific approach to managerial problems involving quantitative factors. This course is designed for managers who wish effectively utilise the tools and techniques of the field of operations research for managerial decision making. Owing to the managerial nature of the course higher emphasis is placed on input requires and interpretation of results as opposed to theoretical mathematical underpinnings. Spreadsheets are extensively used to accomplish the mathematical manipulations. |
Slides, Sheets & Other Shenanigans...
Watch this space for session slide, excel worksheets and useful articles.
|
Wraps & Scribbles
Session 1: Well begun is half done!
That was a fabulous start wasn't it! During this session, we understood what OR and MS are, how they are related and how they are different. We also covered the history of OR and looked at the variety of cases of applications of OR/MS techniques. Finally, we understood what a model is and the steps involved in model building.
For the next session, kindly log in from your laptops with the solver add in installed in MS-Excel by following the instructions from this video. Also, here are some interesting videos on the origins of OR and its various applications for the operations research society. Please do view them!
Session 2 : Baby steps
In our second session, we started off with some rudimentary mathematical models starting with the breakeven point model and tried out the goal seek functionality on MS-Excel which is kind of a pre-cursor to Solver. Next we formulated our first every linear program in the context of a production mix problem. We understand the various aspects of a linear program, decision variables, objective function and constraints. Finally, we looked at graphical solution to the problem and understood the significance of corner points. Discussion on the significance of Binding constraints is pending. Session 3: the path to Excel!
In a session that happened in two parts over two days due to technical difficulties. In this session, through a graphical demonstration, we intuitively understood why the solution of a LP problem always lies in one of its corner points. Further, we also tried our hands on Excel solver for the first time and used it to solve the IMToys problem. IMTrucks and ImMediaTe advertising problems are to be solved for homework.
Here is the link to the online graphing tool I used. You may use it to visualise the solutions to the other problems as well! Link to graph Session 4&5 : Formulation pros...
In these two sessions held over a weekend, we have made the journey from problem formulation to problem solution to effective spreadsheet modelling. We also understood the various categories of linear programming there are and the nuances of how to model these on a spreadsheet. During the second session, I demonstrated how to build a good spreadsheet model with good use of color-coding and range naming etc. The demonstration was mostly a success but for a minor problem that the correct solution couldn't be arrived it. I later realised I had used the wrong inequality sign for one set of constraints. Final solutions have been shared with you over MS Teams. We also looked at the formulations of a couple of blending problems and started discussion on marketing applications. Session 6 : Quiz and Sigma
This week's session was eventful with the first official quiz of the course being held. The wide variation in the performance points to a need on the part of the students to take the opportunity to take a step back and review the happenings of the course so far. Further, we started off with the formulation of fairly large problems and understood the role of expressions using summations in simplifying the notational exposition of complex linear programming problems. |
Session 7 & 8 : Sensitivity Analysis and network take off
To understand the impact of changes in problem parameters on the optimal solution, sensitivity analysis is used. Sensitivity analysis helps us deal with the possibility of the data collected might not be fully accurate and might operate within a range.
Discussion on Transportation problem opened us up to an incredibly useful family of optimization problems. Useful video on Sensitivity Analysis Session 9 & 10 : Untangling the Networks
We discussed three different network optimisation problems: Minimum cost flow, shortest path and maximal flow problems. We used the Aiding allies case from the textbook to understand the how the three problems could be applied.
The larger context network problems and how other problems could transform into easily solvable network problems. Useful videos on MCFP, Maximal flow and Shortest path problems. Session 11 : Integers and complexity!
Implementing integer programming in excel is not complicated. It is just a matter of adding additional integer/binary constraints with a few clicks of a button. However, understanding the ways of crafting integer constraints helps significantly widen the scope of problems that could be solved. This could help produce realistic and implementation friendly results. In this session, we covered the fixed charge, set covering, knapsack and facility location problems. Excel solutions of these problems were demonstrated.
Session 12 : OR in real world!
Practitioner Mr. Prateep Aitha spoke about applications of Operations Research in the Industry. He spoke about the the type of decisions that are appropriate for OR modelling. He also spoke how the underlying OR model needs to be wrapped into a robust decision support system that could aid managerial decision making. Session 13 : Real life is often Non-Linear
During this session, we looked at the challenges in usage of OR in contexts where the objectives are non-linear in nature. We covered the concept of local and global optima and how excel solver solution is not necessarily correct all the time. We also solved a portfolio optimization problem with a non-linear objective. We also looked at the concept of separable programming involving piecewise linear objective function.
Here is an interesting video of how local and global optima work. |