Machine Learning

Noise - Part III - Noise in Predictive Judgments

Noise - Part III - Noise in Predictive Judgments

Translations: RU

How to measure the noise in our daily decisions?

Which is better - expert judgment or mechanical prediction using very simple models/rules? Surprisingly, the model of you beats you! Even a simple regression model based on your past decisions!

How to make rules/models even less noisy? How can rules/models become more biased sometimes?

Why don’t people always use models (if they are so good) and still rely on their bad judgment?

When do we not need models?

State of AI Report 2021

State of AI Report 2021

I just faced with the “State of AI ’2021” report and want to share it. The authors are experts who analyze AI trends, make predictions, and invest(!) in AI. Last year, 5.5 out of 8 their predictions came true. The report has 188(!) pages. What I can highlight briefly: for data scientists & IT managers: Transformers/LLMs expanded far beyond NLP, they will replace recurrent networks JAX framework worth trying for investors & founders: AI-first is now everywhere: new researches in biology, skyrocketing new companies for IPOs and new markets Semiconductor-related companies continue to accelerate and consolidate AI influences politics: army, safety & regulations Here is the link: https://www.
Big Ideas 2021 by ARK invest

Big Ideas 2021 by ARK invest

I’ve just read recently published Big Ideas 2021 by ARK invest (these guys are investing billions of USD into disruptive tech companies) and want to share it. I really recommend to download and overview their slides for everybody who works in high-tech (yes, they are probably over-optimistic because of their role, but it’s really worth to look at): https://drive.google.com/file/d/1z3LpfrAQLRM07JDTi1ZnHEHPeszgIp7O/view?usp=sharing I am especially excited by what they call “Software 2.0” (deep learning).
Travel Industry B2B/B2C solution

Travel Industry B2B/B2C solution

🚆✈️🚌🛳 + Math + Highload + Machine Learning… Two and a half years ago we’ve got a request from the Travel-business IT company, whose B2C IT solution is the basis for one of the world’s largest travel companies that serves over 1 billion passengers every year. What did they want to achieve? They wanted to quickly design and implement an “internal startup” - an innovative B2B/B2C solution for the travel industry.