Asif Ihtemadul Haque

Optimizing Future Environmental Embedding Representations for Crop Performance

Asif Haque / Ongoing research

This work explores how to learn and optimize representations of future environmental conditions, such as climate and soil signals, so that machine learning models can more accurately predict crop performance. The goal is to design embeddings that generalize across regions and seasons, improving forecasting under shifting conditions and supporting data-driven agricultural decisions.

• Machine Learning
• AI
• Environmental Embeddings
• Crop Prediction

Research updates and publications

Asif Haque / In progress

A short space for thesis chapters, preprints, datasets, code, and research notes as they become available.

• Publications
• Thesis
• Preprints

Crop Embedding Toolkit

A research toolkit for training and evaluating environmental embeddings used in crop performance prediction, with reproducible experiment configs.

Python / PyTorch / Pandas / scikit-learn

• Research
• Machine Learning

Environmental Data Pipeline

An ETL pipeline that ingests, cleans, and aligns multi-source environmental and yield data into analysis-ready datasets.

Python / DuckDB / Polars / Airflow

• Data Engineering
• Pipelines

Model Results Dashboard

A lightweight web dashboard for visualizing model metrics, predictions, and embedding spaces during research iterations.

React / TypeScript / Vite / Tailwind CSS

• Web App
• Data Viz

Experiment Tracker CLI

A small command-line tool to log, compare, and reproduce ML experiments with deterministic seeds and config snapshots.

Python / Typer / SQLite

• CLI
• Tooling

Notes on environmental embeddings

A short research note on representing future climate, soil, and seasonal signals for crop performance models.

• Research
• Machine Learning
• Agriculture

Reproducible ML workflow checklist

A compact checklist for keeping machine learning experiments comparable across datasets, regions, and model versions.

• Engineering
• Tooling
• ML