[MUSIC] Hello and welcome to this specialization on analytics for decision making. Im Saumya Sen and I'm an Associate Professor of Information and Decision Sciences at the Carlson School of Management of the University of Minnesota. I'll be teaching courses two and three on optimization and advanced optimization for decision making. I'm here with my colleagues to talk about this specialization. I'd like to invite Professor De Liu to talk about his course. >> Hi, my name is De Liu I'm a Professor of Information and Decision Sciences at Carson School Management University, Minnesota. I'll be glad to teach the first course introduction to predict modeling. Next, I like to introduce Alok. >> Hello, my name is Alok Gupta. I'm Associate Dean of Faculty and Research at Carlson School of Management at University of Minnesota. And I'm also Professor at Information and Decision Sciences Department. I'll be teaching the last course in this specialization, which focuses on simulation modeling. Simulation is one of the most flexible modeling techniques in analytics, and I'm excited to be part of this journey with you. I'll now send it back to Soumya to talk a little bit more about the specialization. >> The field of data analytics has many different methods that one can use, but the first step is to have a clean data set. We're going to use the principles of data engineering in Professor De Liu course to create a clean data set that we can work with. Next we have different methods to choose from. The first is descriptive analytics, which is usually used to gain some initial insights, explore the data and find some patterns. There is predictive analytics which we used to predict the values of variables of interest into the future. For example, you may be interested in predicting costs or capacity of production, or even things like the yield of crops in the future. Then we have causal inference or causal analytics that is used to establish a causal relationship between some variables or factors of interest and an outcome of interest. And Lastly, we have prescriptive analytics in which we use methods such as optimization and simulation in order to provide some best or optimal strategies that the company or the firm should adopt. These techniques usually build on predictive analytics. For example, predictive analytics can be used to project the demands or capacity or cost. And once we have those projected values in the future, you can use it to decide what is the best course of action through using optimization or simulation models. We are going to look at these in the four courses that constitute the specialization. Now I'd like to invite Professor De Liu to talk about his course. The first course in this specialization. >> Thank you Soumya. Predictive modeling is arguably one of the most popular form of analytics. You can use it to predict crop yield using rainfall and temperature or to predict house prices using location, size and other attributes. Or predict Walmart store sales using past historical data. This course is designed to help you get started with predictive modeling. We're going to 1st introduce to you the important steps, concept, and techniques of modeling and then using regression based models as example. Then we move on to teach you how to prepare data for predictive modeling in Excel, including thing with missing values and dealing with different type of data. Finally, we'll spend some time on special and very useful type of predictive modeling time series forecasting. And I hope this course can not only expose you to the art of sciences or predict modeling, but also provide a good exercise for Excel, muscles and get you ready for the remaining courses in this specialization and beyond. Next, I let Soumya talk about his next course. >> Thank you De. So in the course on optimization for decision making, we're going to take the predictions to the next level where we're going to go from predictions to decision making using optimization in phase of different types of organizational constraints, such as budgetary constraints. So we're going to look at different business context that can be modeled in an optimization framework. And we're going to learn how to model linear optimization problem and use simplex method to find the optimal solution. And for doing this we will be using an Excel solver tool and working on several problems of practical relevance. We're going to use the solver tool in Excel to find the best strategy and prescribe that for the particular company. So this is going to provide you with the basics of optimization. In the third course, which is the course on advanced models for decision making, we're going to start looking at more complex models by formulating them again as an optimization problem and then using that to prescribe decision making in very real world context. We're going to see applications of that in finance and cash flow management, in supply chains, and inventory management. In human resource management and questions such as allocation of offices or office assignment and scheduling staff. And we're also going to see applications of these optimization models in production optimization. In all of these, we are going to use Excel, which is an easy to use tool in order to run these optimization problems using the solver built in tool. Next, I'd like to invite Professor Alok Gupta to talk about his course, which is the final course in this specialization. >> Thank you Saumya. As I mentioned earlier, simulation is one of the most flexible analytics tool. The key difference as a tool as compared to other approaches that you learn in this specialization, it is that instead of single answer we can typically get a range of answers with some measure of confidence associated with each outcome. In fact, simulation is often thought of as an art as well as science. And will try to cover both aspects in this course. Typically art refers to the issue of creating a model at the right level and of right type depending upon the questions that you want to answer. I'll also include some tips and tricks that you usually don't find in books or academic literature. And will also use material that you learn in other courses, such as optimization in this specialization to estimate parameters that we need to use in simulation model. I'll use one primary context to keep the continuity of ideas and address building increasingly complex and more sophisticated models. But the principles are general and are applicable to any simulation exercise. Similar to other courses in this specialization will use Excel the primary tool. So large part of the course focuses on what we call Monte Carlo simulation. But we also touched upon discrete event simulation, which goes into details of modeling individual level processes. >> We're really excited to have you as a part of this journey through these courses. While the content of the course has academic rigor, is important to emphasize that you can readily apply these techniques of predictive analytics. Prescriptive analytics, including optimization and simulation in your everyday work environment and turn yourself into a data analytics star. We look forward to seeing you in the course.
