This project aims to develop predictive models for classifying dresses as recommended (1) or not recommended (0) based on their attributes. Using the Dresses Attribute Sales dataset from the UCI Machine Learning Repository, we analyze various categorical features—such as style, price, size, season, neckline, sleeve length, waistline, material, fabric type, decoration, and pattern type—along with the dress rating (0-5) to predict the recommendation outcome.
Steps done as below :
Data Preprocessing: Handling categorical variables and missing data
Exploratory Data Analysis (EDA): Understanding feature importance and distribution
Machine Learning Models: Implementing classification models like Logistic Regression, Decision Trees, Random Forest, and more
Performance Evaluation: Using metrics such as Accuracy, Precision, Recall, and F1-score to assess model performance