学术相关
徐亦达老师的机器学习讲义
- 统计机器学习
- 优化算法
- 概率推断
- 贝叶斯非参数
- 深度学习
- 微分方程
- 机器学习理论
- 强化学习
- 计算机视觉
业界动态
中金丨基于Loop Engineering的自动化因子发现引擎
技术技巧
Kaggle丨Kaggriculture虚拟农场 AI 智能体
Generative-ABM的方向，这个领域非常值得关注
This simulation competition is a turn-based farming game where two players compete on separate farms to see who can earn the most profit by the end of a 30-day season (720 turns).
Your agent acts as themain farmer and can strategically hire farm hands to scale up operations. To succeed, your agent must:
Plant, water, fertilize, and harvest a variety of crops.
Buy, feed, and care for animals to produce eggs, milk, and wool.
Collect and utilize fertilizer to boost crop yields.
Buy neighboring quadrants of land to expand your farm’s footprint.
Trade smart on a dynamic market where prices react to your sales and town demand.
Kaggriculture represents a highly complex environment that models the exact same dynamics found in real-world supply chains, dynamic market pricing, and industrial resource allocation under uncertainty. Underlying mechanics like scheduling resources, optimizing labor, adjusting to supply/demand price changes, and making long-horizon capital investments, serve as a high-fidelity sandbox for training AI to solve complex enterprise operations.
