Data Science people in industries have the ability or potential to command Data Science, but they often lack the rigors that the field requires: The lack of experiences is the general reason, but the lack of the “right” experiences is the key: Most of them do not have either Computer Science or Research background to begin with, this puts them in great disadvantage as the rigors I’m about to discuss are vital and yet easily spotted by a more experienced technical personnel, you will see the Why in just a bit:

One of the biggest human biases in science is…

Time-series problems or problems with numerical targets are often related to Regression, that’s how many have been taught. A recent experience had me considered “retiring regression”, this is certainly an overstatement, but it is worth the thinking because, as one of the oldest and most commonly used techniques, Regression has some fundamental flaws for its use. Here are TWO commonly met problems in Regression:

**1. Regressors are terrible at extrapolation.**

The name “regression” tells a story that relational data points tend to lean towards their expectation, and their velocity of doing so depends very much on how extreme the previous…

Classical computer works by closing and opening the logic gates to signal 0’s and 1’s. When transistors get too tiny, some Quantum properties get in the way. For example, sometimes particles can jump over a closed gate and disrupt the information flow, this is called “Quantum Tunneling”. Quantum Computer seems to be the next step to improve our Computing hardware (given how compute-demanding AI has become), but its counterintuitive properties also offer many interesting directions to improve SOTA in AI. In this blog, we will discuss its significance to Reinforcement Learning on an intuitive level.

Let’s give a short intro…

**Background**: Given the initial state of a dynamical system (“dynamic” means a state occurs based on some probabilities summed up to 1 due to unknown or uncontrollable factors) moves along **Δ**t, we want to predict a trajectory of dynamic states using a model. This model, be it a neural network, serves as an estimator of the dynamic system such that we don’t have to know the underlying dynamics of the system to make meaningful use of it. Some research has found that neural networks are powerful enough to even predict “chaotic systems”(e.g., …

**Context**: For the longest time, I’ve avoided offers in Quantitative Finance for I believe it does not address the heart of people’s problems; however, I do think that there is value in using Quantitative Finance as a problem to develop better utility that does just that. This is because working on algorithmic trading problem can help us better harness Machine’s creativity in coming up with strategy. Programs like AlphaGo & AlphaGoZero are truly creative, not only because they could invent new moves, but also due to the fact that they can be modeled as Generative Models, allowing us to see…

Focusing on RL/Generative Models and beyond. This is a medium for blogging in AI & Compute and networking.