Pinned Repositories
Alert.AI
Created @ 2021 MIT Blueprint Hackathon
baidu-stock-pitch-ml
Supporting statistical analyses, price forecasting, and economic research for Baidu (BIDU) stock pitch
communication-efficient-federated-learning
Quantization framework for reducing communication overhead in federated learning by converting client models from 32-bit floating-point to low-bit formats
neural-style-transfer
AI Art Generator: Implements a neural style transfer optimization algorithm to synthesize novel artistic images from any 2 content and style images
Proximity
HackMIT 2023
python-bootcamp
Class materials for my free 6-week Python bootcamp
real-time-object-detection
Prototype of Real-Time Object Detection System for Self-Driving Cars
spotify-data-analytics
Exploring Music Trends: Machine Learning and Visualizations with Spotify Data
stockscrape
Web scraping for personalized stock tracking
transformer-from-scratch
Implemented and optimized a Transformer network using custom encoder/decoder layers
justinliu23's Repositories
justinliu23/communication-efficient-federated-learning
Quantization framework for reducing communication overhead in federated learning by converting client models from 32-bit floating-point to low-bit formats
justinliu23/facial-recognition-system
Solving the 1:K matching problem with one-shot learning and the triplet loss function
justinliu23/real-time-object-detection
Prototype of Real-Time Object Detection System for Self-Driving Cars
justinliu23/transformer-from-scratch
Implemented and optimized a Transformer network using custom encoder/decoder layers
justinliu23/adam-optimization-algorithm
Combined ideas from RMSProp and Momentum to implement the Adam algorithm from scratch
justinliu23/CARLA-semantic-image-segmentation
Custom U-Net to predict labels for image pixels on the CARLA self-driving car dataset
justinliu23/character-level-text-generation-lstm
A character-level text generation language model to write Shakespearean poems with long-term dependencies from given keywords
justinliu23/convolutional-model-step-by-step
Building a convolutional neural network from scratch
justinliu23/coupon-collector-problem
Mathematical analysis and simulation of the Coupon Collector's Problem
justinliu23/deep-neural-network-step-by-step
Implemented all the building blocks of a neural network to build a deep ANN from scratch
justinliu23/denoising-autoencoder-from-scratch
End-to-end PyTorch implementation of a custom denoising autoencoder for MNIST, engineered to reconstruct clean images from noisy input
justinliu23/efficient-cnn-filter-pruning
Optimizing model efficiency by selectively removing redundant filters while maintaining high classification accuracy on large-scale datasets in PyTorch
justinliu23/federated-learning-simulation
Federated learning simulation with configurable IID and non-IID data distributions
justinliu23/fraud-detection-gradient-checking
Gradient checking to check for errors in implementation of a fraud detection model
justinliu23/goalkeeper-position-recommender
Position recommender for goalkeepers in football games
justinliu23/gradient-descent-initialization-methods
Exploration of various weight initialization methods for gradient descent
justinliu23/gradient-inversion-defense-fl
Evaluation of gradient inversion defenses—pruning, quantization, and noise injection—to protect data privacy in federated learning systems against adversarial reconstruction attacks
justinliu23/justinliu23
justinliu23/logistic-regression-as-neural-network
Building a logistic regression classifier through a neural network mindset
justinliu23/LoRA-parameter-efficient-fine-tuning
A PyTorch-based implementation of LoRA fine-tuning for efficient neural network adaptation
justinliu23/LSTM-music-generator
An LSTM to generate original Jazz music!
justinliu23/lstm-sentiment-classifier
Leveraged the GloVe algorithm to compute word embeddings coupled with a deep LSTM RNN to perform sentiment analysis
justinliu23/MNIST-post-training-quantization
Implementation of post-training quantization and dequantization for neural network weight optimization
justinliu23/mobilenetv2-transfer-learning
Adapting a pretrained model to new data and training a classifier using the Functional API and MobileNetV2
justinliu23/mood-classifier
Created a mood classifier using the TF Keras Sequential API
justinliu23/planar-data-classification
Planar data classification with one hidden layer
justinliu23/resnet50-from-scratch
Implementing the basic building blocks of ResNets in a deep neural network using Keras
justinliu23/rnn-from-scratch
Developed an RNN from scratch, including forward and backward propagation for both basic RNN and LSTM architectures
justinliu23/stable-diffusion-partitioned-controller
Personalized Stable Diffusion featuring text-guided horizontal and vertical partition controllers to apply distinct prompts across specific image regions
justinliu23/word-vector-gender-debiasing
Modifying word embeddings to reduce their gender bias