IEEE

When Multimodal Large Language Models Meet Computer Vision: Progressive GPT Fine-Tuning and Stress Testing

Abstract: The rapid evolution of Multimodal Large Language Models (LLMs) has redefined the landscape of artificial intelligence, with OpenAI’s GPT-4o representing a transformative leap in multimodal ...
note

From a Non-CS Grad Student to a Financial Data Scientist | Self-Introduction

I also learned about F1 scores as evaluation metrics and Optuna for hyperparameter tuning. Also, unrelated to job hunting, I participated in Kaggle after a friend assigned me to a team. It was a text ...
GitHub

CRAFT-Therapeutics/Targeting-HsDHODH-Shape-and-Machine-learning-Guided-Discovery

Curated SDF Dataset (1,792 molecules) │ [A] Molecular Fingerprint Generation ECFP (diam 2/4/6) · FCFP · MACCS Keys │ [B] Baseline ML Classification (PyCaret) LR · RF · ET · XGB · SVM · KNN · NB │ [C] ...
IEEE

An Intelligent Hybrid Model of CNN-BiLSTM for Fault Classification in AC Microgrid

Abstract: The reliable and fast fault classification is paramount for the stability and resilience of your modern power systems as the networks are increasingly dynamic through adopting renewables.
GitHub

Student-depression-lifestyle-Machine-learning-project

The Student Depression Prediction System is an end-to-end Machine Learning and MLOps project developed to predict depression risk among students based on academic, lifestyle, and personal factors. The ...
Frontiers

Integrating Biparametric MRI Radiomics with Clinical Variables Improves PreTreatment Prediction of Prostate Cancer Recurrence

A stacked ensemble model (Random Forest and regularized Logistic Regression as base model; Logistic Regression as meta-model) was developed using five-fold stratified cross-validation, SMOTE balancing ...
Frontiers

Planning cities for extreme heat: a methodological framework for urban resilience in residential building blocks

To maximize predictive fidelity, a Bayesian Optimization (BO) framework governs hyperparameter tuning (see Figure 4). Unlike exhaustive grid searches, BO treats tuning as the global optimization of a ...