Data Science & ML Claude 技能

Transfer Learning Model Adapter

Automates the fine-tuning and adaptation of pre-trained machine learning models for custom datasets and specialized tasks.

Lyra AI Prompt Optimizer

Transforms vague natural language inputs into precision-engineered prompts for Claude, ChatGPT, Gemini, and other large language models.

LLM Infrastructure & Evals

Performs rigorous audits of LLM-powered features to ensure model currency, prompt quality, eval coverage, and observability.

Lumen AI Data Explorer

Enables natural language data exploration, automated SQL generation, and interactive visualization using agent-based AI frameworks.

Advanced Data Rendering (HoloViz)

Optimizes high-performance rendering for datasets exceeding 100M points using Datashader and HoloViews.

Astropy Fundamentals

Processes astronomical data using Astropy for coordinate transformations, FITS file I/O, physical units, and precise time handling.

HoloViz Colormaps & Styling

Masters perceptual color management and accessible visual styling for scientific visualizations using Colorcet and the HoloViz ecosystem.

Pixi Package Manager

Manages scientific Python dependencies and reproducible environments using the pixi package manager with unified conda and PyPI support.

Xarray Multidimensional Data Analysis

Analyzes and manipulates labeled multidimensional scientific datasets using the Xarray library and its ecosystem.

HoloViz Plotting Fundamentals

Generates interactive, publication-quality data visualizations using the HoloViz ecosystem and hvPlot for rapid data analysis.

HoloViz Param & Reactivity

Builds type-safe, declarative configuration systems and reactive Python applications using the HoloViz Param library.

Scientific Python Packaging

Builds and publishes standards-compliant scientific Python packages using pyproject.toml and Hatchling.

Geospatial Visualization

Builds interactive geographic maps and performs spatial data analysis using GeoViews and GeoPandas.

HoloViz Data Visualization

Simplifies the creation of complex, interactive, and multi-dimensional data visualizations using the HoloViz ecosystem.

ATAC-seq Footprint Analysis

Performs transcription factor footprinting and differential binding detection on ATAC-seq data using the TOBIAS framework.

Hi-C Loop Calling

Identifies and calls chromatin loops from Hi-C data files in .mcool, .cool, or .hic formats for genomic visualization and analysis.

Incentive-Based Prompt Engineering

Boosts AI response quality by up to 115% using research-backed techniques like persona assignment, stakes language, and step-by-step reasoning.

Track Generation

Generates normalized BigWig signal tracks from BAM files for ATAC-seq and ChIP-seq visualization.

Known Motif Enrichment Analysis

Identifies enriched transcription factor binding motifs in genomic regions or gene lists using the HOMER bioinformatics suite.

Regulatory Community Analysis for ChIA-PET

Analyzes protein-mediated chromatin interactions to identify and visualize regulatory communities from ChIA-PET datasets.

Functional Enrichment Analysis

Performs Gene Ontology and KEGG pathway enrichment analysis from genomic regions or gene lists with automated R-based visualizations.

ChIP-seq Biological Quality Control

Performs specialized biological validation for ChIP-seq data by calculating cross-correlation metrics and fraction of reads in peaks.

Ramanujan Expander

Analyzes and constructs Ramanujan graphs to achieve optimal spectral expansion and network efficiency within graph-based systems.

Infinity Operads

Manages complex n-ary skill interactions and topos unification using dendroidal Segal spaces and lazy ACSet materialization.

Agent-o-rama Cognitive Learning

Extracts behavioral patterns and trains learning agents for cognitive surrogate systems using temporal and topic analysis.

Entropy Sequencer

Optimizes interaction sequences using information theory and active inference to maximize learning efficiency and information gain.

Scientific Research Automation

Automates the end-to-end scientific research lifecycle from initial data analysis and hypothesis generation to the production of publication-ready LaTeX manuscripts.

AnnData Scientific Data Management

Manages annotated data matrices for single-cell genomics and large-scale biological datasets using the Python AnnData framework.

Cybernetic Immune System

Implements a cognitive immune framework using active inference and information geometry for robust system self-maintenance.

RAG Implementation for LLMs

Builds robust Retrieval-Augmented Generation (RAG) systems using vector databases and semantic search to ground AI responses in proprietary data.