This course provides a comprehensive introduction to regression analysis, using Excel as a tool. The course takes the participants through the principles of regressions, linear curve fitting, the principle of “ordinary least squares, forecasting and develops examples using Excel and Visual Basic for Applications (VBA).

• What are Supervised and Unsupervised learning tasks, and what is regression ?
• The basic linear regression functional form
• Linear Interpolation and linear regression : are they the same ?
• Relationship between variables : correlation and multicollinearity
• Computing the parameters of single variable linear regression:  minimisation of square of residuals
• Practical examples: GDP vs. Internet penetration, material stress/strain data, house prices.
• Excel Analysis tool-pak vs. Excel functions SLOPE, INTERCEPT, FORECAST
• How good is your model ? (assessing quality of fit) : coefficient of determination and distribution of residuals
• Data problems affecting the linear regression outcome: the need for data scaling, overfitting/underfitting, and regression sensitivity to outliers.
• Advanced regression and parametric regression: regression on the coefficients
• Regression on categories : the logistic regression: What is logistic regression and how does it differ from linear regression
• What are the odds?
• Logistic regression functional form
• Calculation example

• Understand what regression means and what it is used for
• Understand the basic mathematical principles behind ordinary least squares and sum of squares of residuals minimization
• Understand the differences and relation between coefficient of determination, coefficient of correlation and residuals
• Interpret properly the output of a linear regression and its coefficients
• Use excel SOLVER and GOALSEEK for regression tasks

Classroom delivery coupled with hands-on applications of classroom knowledge via use cases and interactive sessions.