COMPETITION CASE
Intelligent Assistant for Analyst——Interpretable investing Strategy
When an AI program has enough data for training, it will quickly determine whether a strategy is correct and give advice, while the investor who uses the strategy has no way of knowing the logic of its prediction, which is known as the “algorithmic black box” of machine learning. This competition this year focuses on risk control in the securities market and is titled “Interpretable Investing Strategy”, which solves the black box problem with an algorithmic system that can explain investment strategies. The participating teams need to analyze the relevant data and strategies by using the data and requirements provided by the Organizing Committee of the Competition, develop a risk management platform to provide decision support for risk analysts, focusing on exploring the conduction paths, risk clues, and quantifying the intensity of risk conduction. To this end, the competition organizing committee provides teams with sample data such as processed basic corporate information and its interrelationships, risk scores (macro risk, industry risk, media risk, bond default), corporate blacklists, and so on.
竞赛题目
分析师的智能助手——可解释的投资策略
当AI智能程序拥有足够多的数据训练后，它将迅速判断策略是否正确并给出建议，而使用这些策略的投资者却无法知道其进行预测的逻辑，这被称为机器学习的“算法黑箱"。本届大赛聚焦证券市场的风险控制，赛题为“Interpretable Investing Strategy”，以可解释投资策路的算法系统解决黑箱问题。参赛队伍需要通过使用大赛组委会提供的数据和要求，对相关数据和策略进行分析，开发风险管理平台，为风险分析师提供决策支持，重点探索传导路径、风险线索、量化风险传导强度等内容。为此，大赛组委会为参赛队伍提供了经过处理的企业基本信息及其相互关系、风险评分(宏观风险、行业风险、媒体风险、债券违约)、企业黑名单等样本数据。
COMPETITION REVIEW
