LLM for narrative detection in social media, news, blogs, etc. using NLP, graph science, and clustering with 89.02% accuracy

Outlier prediction and fraud detection using anomaly detection models and Gaussian distribution with high F1-score

Movie/music recommendation using content-based algorithm for user rating prediction, and collaborative filtering for genre and user rating predictions with 97.43% accuracy

Hand-written digits recognition using one-vs-all classification by training regularized multi-class logistic regression classifiers and advanced optimization algorithm (fmincg) with 95.06% accuracy

Facial recognition using K-means clustering and PCA for dimensionality reduction with 98% variations retained

Hand-written numbers recognition using regularized Convolutional Neural Network (CNN) model with 97.52% accuracy

Malignant tumor prediction using multi-class classification with regularization model with 95.18% accuracy

House price prediction using multivariate linear regression with gradient descent model, and normal equation model with 95.22% and 96.4% accuracy respectively

NLP/Network Graph

Trending topics detection in social posts, using NLP, post engagements, graph science, and clustering with 87% accuracy

Prediction on microchips QA inspection based on test scores using binary classification model with feature mapping to create higher-dimension features vector to avoid underfitting, and regularization to avoid overfitting with 98.63% accuracy

Email spam classifier using SVM with Gaussian kernel resulting in a prediction model with 98.70% accuracy

Prediction if students will get admitted to a university based on admission scores using regularized logistic regression model with advanced optimization algorithm (fminunc) with 96.35% accuracy