Dataset: https://www.kaggle.com/c/digit-recognizer/data
Problem statement: Handwritten digit recognization (0-9)
Type of problem: Classification, Supervised
Data type: Images stored in csv file
Performance Measures: Accuracy, Precision, Recall, Confusion Matrix
Feature Importance: Not required
Steps Performed are as follows:
- Importing the necessary libraries
- Loading the Dataset
- Visualizing the Dataset
- Checking whether the Data is balanced or imbalanced
- Performing PCA and t-SNE
- Model building and training
- Conclusion The best model is Random Forest with 96.4% accuracy.