Publications by T. Moudiki
A plethora of datasets at your fingertips
Starting with mlsauce’s next release (v0.9.0, for Python and R), you’ll be able to download a plethora of datasets for your statistical/machine learning experiments (this is a work in progress, it will done from a GitHub branch today). These datasets come from the R-universe, and you’ll be able to use them no matter whether you’re working w...
1632 sym Python (4907 sym/9 pcs) 2 img
A classifier that’s very accurate (and deep)
Version v0.15.0 of nnetsauce is now available on your favorite platforms: PyPI, conda and GitHub for Python; R universe and GitHub for R.The changes in this version are mostly related to Automated Machine learning (AutoML):lazy prediction for classification and regression: see this post for more details on the subject (and remember to use pip insta...
2031 sym R (985 sym/6 pcs) 1 tbl
mlsauce version 0.8.10: Statistical/Machine Learning with Python and R
This week, among other things, I’ve been working on updating mlsauce for both Python and R (that’s version 0.8.10 of the package).mlsauce is a package for Statistical/Machine Learning that contains in particular: AdaOpt, a probabilistic classifier which uses nearest neighbors to obtain predictions. Interestingly, with AdaOpt, one neighbor can s...
3015 sym R (15466 sym/43 pcs) 10 img
AutoML in nnetsauce (randomized and quasi-randomized nnetworks) Pt.2: multivariate time series forecasting
Last week, I talked about an AutoML method for regression and classification implemented in Python package nnetsauce. This week, my post is about the same AutoML method, applied this time to multivariate time series (MTS) forecasting.In the examples below, keep in mind that VAR (Vector Autoregression) and VECM (Vector Error Correction Model) foreca...
1035 sym Python (1448 sym/7 pcs) 1 tbl
AutoML in nnetsauce (randomized and quasi-randomized nnetworks)
Content: Installing nnetsauce for Python Classification RegressionDisclaimer: I have no affiliation with the lazypredict project.A few days ago, I stumbled accross a cool Python package called lazypredict. Pretty well-designed, working, and relying on scikit-learn’s design.With lazypredict, you can rapidly have an idea of which scikit-learn model...
1417 sym Python (3023 sym/8 pcs) 1 tbl
Version v0.14.0 of nnetsauce for R and Python
Version v0.14.0 of nnetsauce is now available for R (hopefully a rapid installation) and Python on GitHub, PyPI and conda. It’s been mainly tested on Linux and macOS. For Windows users, you can try to install of course, but if it doesn’t work, please use WSL2.NEWSupdate and align as much as possible with R version (new plotting function for mul...
642 sym Python (2246 sym/2 pcs) 8 img
An infinity of time series forecasting models in nnetsauce (Part 2 with uncertainty quantification)
As I said a few years ago, this is a family of univariate/multivariate time series forecasting models that I was supposed to present at R/Finance 2020 (this post is 100% Python) in Chicago, IL. But the COVID-19 decided differently.The more I thought about it, namely nnetsauce.MTS (still doesn’t have a more glamorous name), the more I thought ‘I...
2684 sym Python (1987 sym/14 pcs) 14 img
(News from) forecasting in Python with ahead (progress bars and plots)
A new Python version of ahead, v0.9.0 is now available on GitHub and PyPI.ahead is a Python and R package for univariate and multivariate time series forecasting, with uncertainty quantification (in particular, simulation-based uncertainty quantification).Here are the new features in v0.9.0:progress bars for possibly long calculations: the bootstra...
1032 sym Python (912 sym/3 pcs) 8 img
Forecasting in Python with ahead
A new Python version of ahead is now available on GitHub and PyPI. ahead is a package (Python and R) for univariate and multivariate time series forecasting, with uncertainty quantification (in particular, simulation-based).Here are the new features:Align with R version (see https://github.com/Techtonique/ahead/blob/main/NEWS.md#version-070 and htt...
1010 sym
Risk-neutralize simulations
A risk-neutral probability (in Quantitative Finance) is a probability measure in which, on average, all the risky assets return the risk-free rate. Risk-neutral probabilities are widely used for the pricing of derivative products. There are other advanced mathematical definitions of a risk-neutral probability that won’t be discussed here. In R pa...
758 sym R (145 sym/1 pcs)