Data Science -Webinar (Session-1)
The basic objective is be to create an appreciation for what one learns when they take up a course in analytics. So it would not be a training in how to do these things (can't be done even in a day, leave aside an hour), but rather what one should be able to do post doing a course in analytics.
1. What can Analytics help us do? (basic overview of what regression, classification, clustering, forecasting techniques help us achieve/ when we use them) - 10 mins
2. General flow of things while modelling (EDA, Feature Engineering, Feature Selection, Test and Training Data, threshold & optimization) - 10 Mins
3. Data Preparation: - 20 mins
3.i. Feature Engineering
3.i.a. Missing Data Imputation
3.i.b. Categorical Variable & Cardinality
3.i.c. Managing Outliers
3.i.d. Feature Scaling/Normalization
3.ii. Feature Selection
3.ii.a. Eliminating Constant, Duplicate features
3.ii.b. Correlation Matrix
3.ii.c. Statistical test - Information Gain, Univariate Test, ROC-AUC
4. Which tools help you to do these things. What can be done on SPSS/Minitab, why Python/R is so popular for doing these. - 5 Mins
5. Q&A - 10 Mins