Simon Ouellette is creating content about stochastic data science
3

patrons

I believe conventional data science and machine learning approaches are inadequate when it comes to solving fundamentally stochastic, non-stationary problems (such as in the world of finance). The vast majority of machine learning techniques expect the data points to be i.i.d., and the training dataset's distribution must match that of the validation set. 

The quants have been addressing this fundamental problem for a long time, with their emphasis on stochastic calculus and on models that typically account for such random variables. However, the classical quantitative finance approach doesn't seem to have caught up with recent data inference techniques and concepts. It is stuck in a top-down, "mathematical derivation"-only mode.

As a result, the main focus in my articles, talks and meetups, is to address these shortcomings by helping bridge the gap between cutting edge data science and stochastic modelling. This is my humble effort at shedding light on an area of data science that is, in my opinion, much neglected and under-developed.

The content I am currently creating is:

Tiers
Supporter
$2 or more per month
Your support is much appreciated
Participant
$5 or more per month
Gain remote live video access to my meetups, talks, livestream Q&As and be able to ask questions and participate in real-time.
I believe conventional data science and machine learning approaches are inadequate when it comes to solving fundamentally stochastic, non-stationary problems (such as in the world of finance). The vast majority of machine learning techniques expect the data points to be i.i.d., and the training dataset's distribution must match that of the validation set. 

The quants have been addressing this fundamental problem for a long time, with their emphasis on stochastic calculus and on models that typically account for such random variables. However, the classical quantitative finance approach doesn't seem to have caught up with recent data inference techniques and concepts. It is stuck in a top-down, "mathematical derivation"-only mode.

As a result, the main focus in my articles, talks and meetups, is to address these shortcomings by helping bridge the gap between cutting edge data science and stochastic modelling. This is my humble effort at shedding light on an area of data science that is, in my opinion, much neglected and under-developed.

The content I am currently creating is:

Recent posts by Simon Ouellette

Tiers
Supporter
$2 or more per month
Your support is much appreciated
Participant
$5 or more per month
Gain remote live video access to my meetups, talks, livestream Q&As and be able to ask questions and participate in real-time.