Data Science

Forecasting: Are ML Models ‘Winning’ Because of Less Preprocessing Needs?

In the world of forecasting there is a bit of mystery. Many expert forecasters continue to use statistical models despite the strong push of machine learning (ML) and deep learning neural network (NN) models. Those less familiar with the world of forecasting assume, because of the dominance of ML/NN models in computer vision and language […]

Forecasting: Are ML Models ‘Winning’ Because of Less Preprocessing Needs? Read More »

A Different Look at Intermittent Demand Forecasting

Demand forecasting is exactly how it sounds: forecasting demand. This can take many forms but the most classic example is predicting how many products will be sold by a retailer on each particular day. This forecasting is in turn used to drive business value in various ways. The most common is better inventory management, but

A Different Look at Intermittent Demand Forecasting Read More »

To Be, or Not to Be AGI, and how do I use it? That is the question

Is ChatGPT and the its family of LLMs alive? If not, will models like these someday (soon) be alive? Do we need to be afraid of them destroying our world as we know it? And regardless of whether these things are alive or not, will they be useful? Those questions are big questions right now,

To Be, or Not to Be AGI, and how do I use it? That is the question Read More »

Winning the M6 Competition: Financial Forecasting Decision Category

For the last couple of years, I have spent a large portion of my life attempting to solve the unsolvable: forecasting time series to predict the future. I have personally thought myself quite clever for the various innovations I have come up with, and the thousands of users on my free and open-source AutoTS project

Winning the M6 Competition: Financial Forecasting Decision Category Read More »

Introducing the Cassandra Model

I have always been particularly impressed with the forecasting performance of the Prophet model. It has only a few parts: seasonality, holidays, any external regressors, and with those it has some of the most consistent forecasting performance around. Also, being a linear model (with Bayesian parameter estimation), Prophet offers a decomposable explainability/interpretability, i.e. Monday has

Introducing the Cassandra Model Read More »