Excavate-25

This contains the models we made for Excavate Composit'25 using Machine Learning modelling techniques

Tech Stack :

Technology & Description :

Technology	Description
	Mathematical Computation
	Data Manipulation and Analysis
	Used for Plotting Graphs
	Regression & Classification
	Deep Learning
	Visualization

Achievements :

1. Regression : Fine-tuned Linear Regression models with One-Hot Encoding & Imputation techniques, achieving an impressive R² = 0.798 for accurate excavation predictions.
2. Classification : Focused on Random Forest, SVM, & K-Means, optimizing the Random Forest model for 90% accuracy, minimizing misclassifications in excavation categorization.
3. Deep Learning Synergy – Leveraged TensorFlow & Keras to integrate ML models, boosting accuracy to 92% with Fully Connected Neural Networks and advanced optimization techniques.