I build computational tools that bridge astrophysics and machine learning — turning telescope data into discoverable patterns. Currently focused on three problems:

Each project feeds the next. Exoplanet habitability teaches me classification. Molecular line ID teaches me spectral analysis. Chemical kinetics teaches me the physics underneath. The goal is a unified tool: observe a star-forming region, identify its molecules, infer its chemistry, predict its planet-forming potential.

• Astrochemistry: Molecular formation and destruction in interstellar environments. Chemical modeling of star-forming regions. ALMA spectral line surveys.
• AI for Astronomy: Machine learning for spectral classification, exoplanet detection, and automated analysis of large astronomical datasets.
• Computational Physics: Numerical solutions to astrophysical ODEs. Chemical kinetics simulations. Orbital mechanics and N-body problems.
• Statistical Methods: Dimensionality reduction (PCA, t-SNE), Bayesian inference, uncertainty quantification in astronomical measurements.

End-to-end pipeline for exoplanet habitability classification.

NASA Exoplanet Archive → Data Cleaning → Feature Engineering 
→ Random Forest / Logistic Regression → Evaluation → Publication Figures

Key technical choices:

• Kasting model for habitable zone boundaries (physics-based, not arbitrary)
• Class balancing because habitable planets are ~1% of the dataset
• Derived features: planetary density, insolation flux, equilibrium temperature
• Reproducible: fixed random seeds, documented data sources, unit tests

What I learned building this:

• How to handle missing data in astronomical datasets (imputation vs. removal)
• Why precision matters more than accuracy for rare-class classification
• How to structure a research codebase (data/raw/, data/processed/, src/, tests/)

Literature reviews, reproduced methodologies, and research proposals.

Current depth:

• Harada et al. (2024) — ALCHEMI survey: PCA classification of 100+ molecular species in NGC 253
• Kasting et al. (1993) — Habitable zone physics: the equations behind my mini project
• Draine (2011) — ISM physics: the environment where astrochemistry happens

What I'm reading next:

• Spectroscopic data cubes: how ALMA turns photons into 3D datasets
• UCLCHEM: time-dependent chemical networks for molecular cloud modeling
• CNN architectures for 1D spectral data (not images — spectra)

This portfolio. Built with pure HTML/CSS/JS. No frameworks. Loads instantly on any device.

Why no React/Vite/Next.js: Any professor or enthusiast opening this on a phone in an airport doesn't need a build step. They need content. The tech stack is invisible by design.

Contains:

• Interactive 3D models (Three.js): habitable zone solar system, water molecule, ALMA spectral window
• MathJax equations: habitable zone physics, insolation flux, rotational transitions
• Plotly visualizations: exoplanet scatter plots, chemical kinetics time evolution
• Timeline: from undergraduate projects to graduate research

Why Japanese: I want to read Japanese astrophysics papers without translation loss and collaborate directly with researchers at NAOJ, ALMA-J, and other institutions without language as a barrier. The grammar of a language shapes the grammar of its science.

Current tools: Bunpro (grammar), Ruupa (kanji radicals), Duolingo (vocabulary), NHK World (listening), Tadoku graded readers (reading).

Current CGPA: 8.4+

Programming: Python, C, Bash. NumPy, SciPy, pandas, scikit-learn, TensorFlow, Keras, Matplotlib, Plotly.

Workflow: Git, GitHub, Jupyter, VS Code, EndeavourOS + KDE Plasma, LaTeX, Markdown.

Astronomy: NASA Exoplanet Archive, ALMA Science Archive, CDMS/JPL spectral line catalogs, UCLCHEM.

Mathematics: Linear algebra, multivariable calculus, ODEs, probability & statistics, discrete mathematics, proof techniques.

Why Linux: Reproducible environments. No hidden dependencies. Everything is a file. The command line is the research interface.

I write science fiction — Echoes of Eternity has reached 110,000+ readers. I believe storytelling and scientific research are the same skill: observing patterns, building coherent worlds, and communicating them clearly. The protagonist is a researcher who discovers that galaxies hold echoes of events that haven't happened yet. The science is real. The struggle is mine.

Not on GitHub. Writing is a separate channel. But it informs how I structure research papers: every equation needs a narrative, every figure needs a story.

I am actively seeking:

• Research collaborations in astrochemistry, radio astronomy, or AI for science
• Mentorship from researchers who've made the transition from undergraduate to graduate research
• Japanese language practice partners (especially science/astronomy context)
• Feedback on my code, papers, or research direction

Email: zaheerares1256@gmail.com

Discord: Zackákro (display name) · retrogradestar (username)

Response time: Usually within 24 hours. Slower during exam periods (November, May).

Preferred contact: Email for formal collaboration, GitHub issues for code discussion, Discord for casual chat.