# Machine learning by Toby Dylan Hocking, PhD

## Table of Contents

## Email: toby.hocking@r-project.org

## Machine learning consulting services

I am an expert in designing and implementing machine learning models for recognizing and predicting patterns in big data sets. I have published several academic papers at the major machine learning conferences (see my Google scholar profile and my list of reproducible papers with accompanying source code).

I am available for hire as a consultant for:

- Designing new machine learning models for your data. Send me your data set along with a description of what you are trying to predict, and I will design a model and write custom code for you.
- Teaching you how to use free/open-source software for machine learning (R/Python/C/C++). Explain to me what kinds of machine learning models and/or software tools you would like to use, and I will create a custom seminar for your particular needs. List of previous teaching about machine learning.

## Machine learning software products

I have written several free/open-source software packages for machine learning.

- clusterpath implements an algorithm for unsupervised hierarchical clustering (R package with C++ code).
- rankSVMcompare implements a supervised learning algorithm for ranking and comparing objects, based on labeled preference pairs (R package).
- WeightedROC can compute a ROC curve for a weighted binary classification problem (R package).
- PeakSegJoint implements a discrete optimization algorithm for supervised peak detection (R package with C code).
- PeakSegOptimal implements linear time algorithms for constrained optimal changepoint detection (R package with C++ code).
- penaltyLearning implements algorithms for supervised learning of penalty functions for optimal changepoint detection (R package with C++ code).

I am available for hire for teaching you about these software packages.