SUK · Third Year B.Tech. Computer Science and Engineering (Data Science) · NEP 2020
Exploratory Data Analysis
DS3205 · Elective III · Semester VI · 3 credits · ISE 10 + mid-semester 30 + ESE 60
One of the Elective III options.
Unit 1: Introduction to EDA
7 hoursEDA: definition, need and steps; datasets, variables and their types, numerical and categorical variables, cardinality; relationships between variables, covariance, correlation, multicollinearity and the normal distribution.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Exploratory data analysis
- Types of data and measurement scales
- Correlation and covariance
Unit 2: Handling Missing Data and Data Encoding
7 hoursInterpreting and imputing missing data (mean, mode, median, min, max, forward and backward fill) or removing it; one-hot, ordinal, label and mean encoding.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Handling missing data
- Encoding categorical variables: one-hot, ordinal and target encoding
Unit 3: Variable Discretisation and Outliers
7 hoursEqual-interval discretisation followed by categorical encoding; interpreting outliers, trimming, capping at arbitrary limits and zero coding.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Binning, binarisation and interaction features
- Outlier detection: z-score and IQR
Unit 4: Feature Scaling
7 hoursWhy scaling matters; normalisation versus standardisation; MaxAbs, robust, quantile-transformer and power-transformer scalers.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Normalisation and standardisation
- Box-Cox and power transforms
Unit 5: Feature Engineering
6 hoursThe curse of dimensionality; feature elimination with PCA and LDA; wrapper and embedded feature selection; multicollinearity and the VIF.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Principal component analysis
- Linear discriminant analysis
- Feature selection: filter, wrapper and embedded methods
Unit 6: Data Balancing
6 hoursInterpreting classification datasets and the impact of imbalance; under- and over-sampling, k-fold cross-validation, SMOTE, the balanced bagging classifier and threshold moving.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Imbalanced classification
- Cross-validation and model evaluation
Unit contents are summarised from the official SUK syllabus . AiCanCode.org is not affiliated with Shivaji University, Kolhapur.