SUK · Third Year B.Tech. Computer Science and Engineering (Artificial Intelligence and Machine Learning) · NEP 2020
Feature Engineering
AIML3202 · Semester VI · 4 credits · ISE 10 + mid-semester 30 + ESE 60
A 4-credit course (3 lectures and 1 tutorial a week).
Unit 1: Introduction to Feature Engineering
6 hoursWhy feature engineering matters; types and levels of data and of features; evaluating ML algorithms and feature-engineering procedures; feature understanding.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Feature engineering
- Types of data and measurement scales
- Data quality dimensions
Unit 2: Basics of Feature Representation
5 hoursScalars, vectors and spaces; binarisation, binning and log transforms; min-max scaling, standardisation and ℓ2 normalisation; feature selection.
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
- Square-root, arcsine, log and logit transforms
- Normalisation and standardisation
Unit 3: Features of Text Data
6 hoursBag-of-words vectors; stop words, frequency filtering and stemming; n-grams and phrases, parsing and tokenisation, collocation extraction; TF-IDF.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Tokenisation, spelling correction and minimum edit distance
- Word representations
- TF-IDF and vector space models
Unit 4: Features of Categorical Data
6 hoursOne-hot, dummy and effect coding and their pros and cons; large categorical variables: feature hashing and bin counting.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Encoding categorical variables: one-hot, ordinal and target encoding
Unit 5: Feature Selection
7 hoursGoals and classes of feature-selection methods; irrelevant features; overfitting to predictors and external validation; simple filters, recursive feature elimination and stepwise selection.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Feature selection: filter, wrapper and embedded methods
Unit 6: Feature Transformation
6 hoursTransformation techniques; the curse of dimensionality and PCA; benefits of dimensionality reduction; feature construction.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Feature extraction and PCA
- Principal component analysis
Unit contents are summarised from the official SUK syllabus . AiCanCode.org is not affiliated with Shivaji University, Kolhapur.