Suggestive AI, a New Paradigm in Artificial Intelligence
December 8th, 2022
Suggestive AI denotes AI models that make contextual suggestions, rather than point predictions. Recent models such as ChatGPT and Stable Diffusion could be described as Suggestive AI models. My prediction is that Suggestive AI models more naturally fit into human driven business processes and eliminate barriers to adoption that previous machine learning and AI models suffered from.
Simultaneous Forecasting of Multiple Time-Series
September 30th, 2019
How can we leverage relations between multiple time-series to improve the accuracy of our forecasts?
Comparison of Machine Learning Methods for Forecasting Web Traffic
August 22nd, 2019
Here I investigate how well several machine learning methods perform on a real world forecasting experiment.
Optimal Inventory Management with Model Predictive Control
August 12th, 2019
How can we optimally manage our inventory with uncertain demand? Turns out, you need to be able to predict the future and then optimize for it.
Creating Artificial Intelligence with Model Predictive Control
July 31st, 2019
There is a lot of hype around reinforcement learning right now. Many of the current approaches use so-called model free methods in which agents learn without really understanding the environment they operate in.
Time Series Forecasting with Harmonic Regression
July 24th, 2019
Using periodic functions and linear regression to forecast.