说到特征降维/选择的问题，大部分EDA的套路都是从model训练的loss来判断feature importance。其实有一个简单易行而且很有效的办法是在CV里面用做feature permutation，对原始特征shuffle得到shadow（也可以加一些噪音），在通过zscore比较两者差异来判断importance，不断遍历筛选。在ESLII中593页有提到这个办法。R里面有一个包Boruta可以做这件事，py也有：https://github.com/scikit-learn-contrib/boruta_py
BY Data Science Archive
Durov said on his Telegram channel today that the two and a half year blockchain and crypto project has been put to sleep. Ironically, after leaving Russia because the government wanted his encryption keys to his social media firm, Durov’s cryptocurrency idea lost steam because of a U.S. court. “The technology we created allowed for an open, free, decentralized exchange of value and ideas. TON had the potential to revolutionize how people store and transfer funds and information,” he wrote on his channel. “Unfortunately, a U.S. court stopped TON from happening.”
Bitcoin is a decentralized digital currency that you can buy, sell and exchange directly, without an intermediary like a bank. Bitcoin’s creator, Satoshi Nakamoto, originally described the need for “an electronic payment system based on cryptographic proof instead of trust.” Each and every Bitcoin transaction that’s ever been made exists on a public ledger accessible to everyone, making transactions hard to reverse and difficult to fake. That’s by design: Core to their decentralized nature, Bitcoins aren’t backed by the government or any issuing institution, and there’s nothing to guarantee their value besides the proof baked in the heart of the system. “The reason why it’s worth money is simply because we, as people, decided it has value—same as gold,” says Anton Mozgovoy, co-founder & CEO of digital financial service company Holyheld.
Data Science Archive from ca