COURSE CONTRIBUTION TO PROGRAM
The objective of this course is to introduce the main concepts of Business Analytics, focusing in particular on the interpretation and validity of the results that are obtained as well as on how to use them to create new business strategies. To this end, we will introduce BigML (www.bigml.com) that will provide us the right platform for statistical and Machine Learning modeling. At the end of the course, students should:
- Be familiar with Data and Business Analytics.
- Be acknowledgeable on how companies are leveraging data techniques to outperform.
- Understand the fundamental statistical concepts applied to business and decision-making.
- Be able to solve a business case applying Business Analytics techniques and Machine Learning models.
By the end of the course, participants are expected to have acquired the knowledge on the foundation of Business Analytics and be capable of analyzing a business problem with data. They should be able to understand, discuss and define business analytics strategies for real business contexts and challenges
Course Learning Objectives
The objective of this course is to introduce the main concepts of Artificial Intelligence, focusing in particular on the basics of Machine Learning to solve business challenges. To this end, we will introduce BigML (www.bigml.com) that will provide us the right platform for Machine Learning modeling. At the end of the course, students should:
- Be familiar with Data and Business Analytics.
- Be acknowledgeable on how companies are leveraging data techniques to outperform.
- Understand the fundamental statistical concepts applied to business and decision-making.
- Be able to solve a business case applying Business Analytics techniques and Machine Learning models.
By the end of the course, participants are expected to have acquired the knowledge on the foundation of Machine Learning applied to Business being capable of analyzing a business problem with data. They should be able to understand, discuss and define business analytics strategies for real business contexts and challenges.
Why is learning about Business Analytics important?
Data Analytics is the next frontier for productivity, innovation and competition. Digitalization is fully taking place in private and public sectors of the economy generating vast amounts of relevant data. Data is the new oil of the economy, and this course is about extracting value from data. Paraphrasing MIT Professors Erik Brynjolfsson and Andrew McAfee, "managers who use artificial intelligence will replace those who are not able to take advantage of the possibilities that it offers.? They were basically referring to the transformation power of Machine Learning that is a subfield of Artificial Intelligence. Business Analytics (BA) has three different angles:
- Descriptive Business Analytics which is used to understand the past and present of an organization through the data it generates. This part of the discipline is strongly based on statistics and data visualization.
- Predictive Business Analytics which allows organizations to make predictions on what will happen based on the collected data on the past. Predictive BA uses Machine Learning (subfield of AI) to create forecasting models.
- Prescriptive Business Analytics that generates recommendations on actions in order to influence the future. In this course, we are not going to explore this area, although there will be some material available describing Recommender Systems.
On the other hand, there is and will be a significant shortage of analytical talent both in technical and business job positions. This course is not about becoming a data scientist but about becoming a manager that deeply understands the potential of data analytics to solve real business needs. BA will become (and it is already) key for any organization that is willing to keep its competitiveness. Therefore, managers need to be able to talk the same language as data scientists. This is a new competence that will be needed in all stages of a company, for all different areas and departments, and in all sectors and industries.
The main objective of this course is to educate yourself to be able to improve decision-making in any discipline and organization by using business analytics. We will learn how to use existing platforms to perform many data analytics tasks without requiring programming skills.
CONTENT
1. About the Course 1. About the course. The first part of the first session is about introducing the topic of the course and its methodology. Evaluation mechanisms are also detailed.
¿ About the course
¿ Evaluation
¿ The tools for the course
¿ Machine Learning. This session is devoted to understanding Machine Learning from a business point of view. General concepts such as supervised vs unsupervised learning will be explored. Within the supervised learning models, we will also see the two types of problems that they can solve: regression and classification. The session introduces BigML as well as a platform to model a problem with a ML model.
i. What is Machine Learning?
ii. Supervised and Unsupervised Machine Learning
iii. Problems that can be solved with supervised ML (Regression, Classification)
|
2. Supervised Learning: Classification 2. Supervised Learning: Classification. In this session, we will study in depth how to solve classification problems with supervised Machine Learning. Also, we will learn how to evaluate a classification model and interpret this evaluation in the business context. Several examples in BigML will be shown in class.
¿ Classification problems examples
¿ Classification models (logistic regression, decision trees, random forests, ensembles)
¿ Evaluation of classification models (confusion matrix, accuracy, precision, recall, etc)
¿ Interpretation of classification evaluations in the business context
¿ Examples in BigML
|
3. Supervised Learning: Regression 3. Supervised Learning: Regression. In this session, we will study in depth how to solve regression problems with supervised Machine Learning. Also, we will learn how to evaluate a regression model and interpret this evaluation in the business context. Several examples in BigML will be shown in class.
¿ Regression problems examples
¿ Regression models (linear regression, decision trees, random forests, ensembles, deep learning)
¿ Evaluation of regression models (MAE, RMAE, R squared)
¿ Interpretation of regression evaluations in the business context
¿ Examples in BigML
|
4. Unsupervised Learning 4. Unsupervised Learning. This session gives an overview of unsupervised learning with emphasis on two type of models: clustering and association rules. It includes real business examples in BigML.
¿ Unsupervised Learning overview with examples.
¿ Clustering modelling. Examples in BigML
¿ Association rules. Examples in BigML
|
Methodology
The course format and methodological approach are based on a combination of theory and practice. During the main sessions, participants will be provided with the material needed to follow the course. The material includes both the theory of the different subjects to be discussed and the data needed to practice with real business case scenarios.
Two group assignments will be proposed to solve business cases using the methods learned in class covering supervised learning both for classification and regression, and unsupervised learning.
Throughout the course, students will work by groups on a final project on the resolution of a business case. The topic is to be chosen by each group. Intermediary deliveries will make sure right progress is done during the course. At the end of the course, a group presentation will describe all the findings of the real case which is part of the evaluation of the course.
