OpenClaw

ml-evolution-agent

Auto-evolving ML competition agent. Learns from each experiment, accumulates HCC multi-layer memory, and continuously improves LB scores. Inspired by MLE-Bench #1 ML-Master methodology.

guohongbin-git0 来源安装数☆ 0 仓库 Stars

查看来源 ↗安装与使用 SKILL.md ↗

安装与使用

请根据来源仓库和你的 AI 工具说明安装此技能。先阅读技能文件和依赖要求，再添加到工作流。

SkillHub CLI

skillhub install guohongbin-git/ml-evolution-agent

打开来源仓库，查看说明和兼容环境后，按来源提供的命令或步骤安装。

将来源链接提供给你的 Agent，让它先读取说明、核对依赖，并向你说明将执行的安装步骤。

查看来源 ↗使用指南

技能说明

ML Evolution Agent 🤖

Auto-evolving ML competition agent that learns from every experiment.

What This Skill Does

Auto-evolves ML models for Kaggle-style competitions
HCC Multi-layer Memory - Episodic, Pattern, Knowledge, Strategic layers
Continuous improvement - Each phase learns from previous failures/successes
Resource-aware - Respects system limits (time, memory, API quotas)

When to Use

User mentions Kaggle competition
Tabular data classification/regression tasks
Need to beat a target LB score
User wants automated ML experimentation

Quick Start

# Initialize
from ml_evolution import MLEvolutionAgent

agent = MLEvolutionAgent(
    competition="playground-series-s6e2",
    target_lb=0.95400,
    data_dir="./data"
)

# Run evolution
agent.evolve(max_phases=10)

HCC Memory Architecture

Layer 1: Episodic Memory
├── Experiment logs (phase, CV, LB, features, params)
├── Success/failure records
└── Resource usage tracking

Layer 2: Pattern Memory
├── What works (success patterns)
├── What fails (failure patterns)
└── When to use each approach

Layer 3: Knowledge Memory
├── Feature engineering techniques
├── Model configurations
├── Hyperparameter knowledge
└── Domain-specific features

Layer 4: Strategic Memory
├── Auto-evolution rules
├── Resource management rules
├── Exploration-exploitation balance
└── Competition-specific strategies

Proven Techniques (from real competitions)

Feature Engineering

Technique	Effect	Best For
Target Statistics	+0.00018 LB	All tabular data
Frequency Encoding	+0.00005 LB	High-cardinality features
Smooth Target Encoding	+0.00003 LB	Prevent overfitting
Medical Indicators	+0.00006 CV	Health data

Model Configurations

Model	Best Params	Weight
CatBoost	iter=1000-1200, lr=0.04-0.05, depth=6-7	50%
XGBoost	n_est=1000-1200, lr=0.04...

来源信息

Source: github
Repository: openclaw/skills
Path: skills/guohongbin-git/ml-evolution-agent
Version: —
Commit: b167e156c9bb654200afdcc2d01ff25298016db9
分类来源: rules
状态: active
可见性: public

这些信息来自收录时的来源分析，可能滞后；使用前请核对最新来源文件。

来源内容

SKILL.md

# ML Evolution Agent 🤖

**Auto-evolving ML competition agent** that learns from every experiment.

## What This Skill Does

1. **Auto-evolves ML models** for Kaggle-style competitions
2. **HCC Multi-layer Memory** - Episodic, Pattern, Knowledge, Strategic layers
3. **Continuous improvement** - Each phase learns from previous failures/successes
4. **Resource-aware** - Respects system limits (time, memory, API quotas)

## When to Use

- User mentions Kaggle competition
- Tabular data classification/regression tasks
- Need to beat a target LB score
- User wants automated ML experimentation

## Quick Start

```python
# Initialize
from ml_evolution import MLEvolutionAgent

agent = MLEvolutionAgent(
    competition="playground-series-s6e2",
    target_lb=0.95400,
    data_dir="./data"
)

# Run evolution
agent.evolve(max_phases=10)
```

## HCC Memory Architecture

```
Layer 1: Episodic Memory
├── Experiment logs (phase, CV, LB, features, params)
├── Success/failure records
└── Resource usage tracking

Layer 2: Pattern Memory
├── What works (success patterns)
├── What fails (failure patterns)
└── When to use each approach

Layer 3: Knowledge Memory
├── Feature engineering techniques
├── Model configurations
├── Hyperparameter knowledge
└── Domain-specific features

Layer 4: Strategic Memory
├── Auto-evolution rules
├── Resource management rules
├── Exploration-exploitation balance
└── Competition-specific strategies
```

## Proven Techniques (from real competitions)

### Feature Engineering
| Technique | Effect | Best For |
|-----------|--------|----------|
| Target Statistics | +0.00018 LB | All tabular data |
| Frequency Encoding | +0.00005 LB | High-cardinality features |
| Smooth Target Encoding | +0.00003 LB | Prevent overfitting |
| Medical Indicators | +0.00006 CV | Health data |

### Model Configurations
| Model | Best Params | Weight |
|-------|-------------|--------|
| CatBoost | iter=1000-1200, lr=0.04-0.05, depth=6-7 | 50% |
| XGBoost | n_est=1000-1200, lr=0.04...

搜索