yujansaya
From crafting code to curating the joys of parenthood, I'm now channeling my creativity into a new adventure—Machine Learning and AI Engineering.
London
Pinned Repositories
a3c_agent_pytorch
This code implements an Asynchronous Advantage Actor-Critic (A3C) algorithm using PyTorch to train an agent to play the Atari game "Boxing"
ai_math_solver
The project deploys a pre-trained language model to efficiently solve mathematical problems by generating textual answers based on input questions. It enhances performance with optimizations like quantization and extracts numerical answers from the generated text.
credit_risk_model
Create a model to predict which clients are more likely to default on their loans.
diarization_sentiment_analyse
A web app for real-time transcription, speaker diarization, and sentiment analysis
gemma
Google – AI Assistants for Data Tasks with Gemma
gemma_prompt_recovery
The project aims to generate prompts for essay rewriting by fine-tuning the Gemma model on a dataset of original and rewritten texts, integrating LoRA for efficient training and inference. Tools: Python, Hugging Face's Transformers and Datasets, Gemma, LoRA, Accelerate, PyTorch, Pandas.
harmful_brain_acitivity
Harmful Brain Activity Detection Classification - Classify seizures and other patterns of harmful brain activity in critically ill patients
house_price_prediction
House Price Prediction (Kaggle)
kaggle_fraud_detection
Credit Card Fraud Detection
molecule_binding_prediction
The project utilises ML models and ensembles to predict molecular binding, leveraging fingerprints and protein features. It evaluates model performance, integrates calibration for refined predictions, and aims to optimise accuracy in chemical compound interactions. Tools: DuckDB, RDKit, XGBoost, CatBoost, LightGBM, Ensemble Learning,Calibration
yujansaya's Repositories
yujansaya/harmful_brain_acitivity
Harmful Brain Activity Detection Classification - Classify seizures and other patterns of harmful brain activity in critically ill patients
yujansaya/a3c_agent_pytorch
This code implements an Asynchronous Advantage Actor-Critic (A3C) algorithm using PyTorch to train an agent to play the Atari game "Boxing"
yujansaya/ai_math_solver
The project deploys a pre-trained language model to efficiently solve mathematical problems by generating textual answers based on input questions. It enhances performance with optimizations like quantization and extracts numerical answers from the generated text.
yujansaya/credit_risk_model
Create a model to predict which clients are more likely to default on their loans.
yujansaya/diarization_sentiment_analyse
A web app for real-time transcription, speaker diarization, and sentiment analysis
yujansaya/gemma
Google – AI Assistants for Data Tasks with Gemma
yujansaya/gemma_prompt_recovery
The project aims to generate prompts for essay rewriting by fine-tuning the Gemma model on a dataset of original and rewritten texts, integrating LoRA for efficient training and inference. Tools: Python, Hugging Face's Transformers and Datasets, Gemma, LoRA, Accelerate, PyTorch, Pandas.
yujansaya/house_price_prediction
House Price Prediction (Kaggle)
yujansaya/kaggle_customer_segmentation
Mall Customer Segmentation Data
yujansaya/kaggle_fraud_detection
Credit Card Fraud Detection
yujansaya/kaggle_titanic
Machine learning to predict which passengers survived the Titanic shipwreck
yujansaya/molecule_binding_prediction
The project utilises ML models and ensembles to predict molecular binding, leveraging fingerprints and protein features. It evaluates model performance, integrates calibration for refined predictions, and aims to optimise accuracy in chemical compound interactions. Tools: DuckDB, RDKit, XGBoost, CatBoost, LightGBM, Ensemble Learning,Calibration
yujansaya/raha_beach_ocr
OCR Technical Assignment
yujansaya/yujansaya