BLOG

A selection of our posts on Towards Data Science, KDnuggets, The Startup and The Data Driven Investor.

  • Solving the Kaggle Telco Customer Churn challenge in minutes with AuDaS
    by Charles Brecque on 17/10/2018 at 14:11

    Solving the Kaggle Telco Customer Churn challenge in minutes using AuDaSAuDaS is the automated Data Scientist developed by Mind Foundry which aims to allow anyone, with or without a background in Data Science to easily build and deploy quality controlled Machine Learning pipelines. AuDaS empowers Business Analysts and Data Scientists by allowing them to easily insert their domain expertise in the model building process and extract actionable insights.In this tutorial we are going to see […]

  • Warm Starting Bayesian Optimization
    by Charles Brecque on 15/10/2018 at 13:31

    (source)Hyper-parameter tuning is required whenever a Machine Learning model is trained on a new data-set. Nevertheless, it is often foregone as it lacks a theoretical framework which I have previously tried to demystify here:Demystifying Hyper-Parameter tuningOne approach which systematises intelligent and efficient hyper-parameter tuning is Bayesian Optimization which builds a probabilistic surrogate of the tunable problem to recommend optimal parameters. It gradually builds up its […]

  • Data Wrangling with Data Report (Part 1/3)
    by Charles Brecque on 12/10/2018 at 13:09

    A useful package for Wrangling large data setsThe story behind Data ReportBig Data does not always equate to quality data but its sheer size and the lack of appropriate tools often prevents us from making that judgement. At Mind Foundry, we built Data Report to efficiently profile large data sets and trim them down through cardinality and correlation analysis. There are many tools out there such as Seaborn and pandas profiling, but we have made Data Report easier to use whilst […]

  • Bayesian Optimization for Quantitative Trading
    by Charles Brecque on 10/10/2018 at 13:44

    In this tutorial we are going to see how Bayesian Optimization can reduce the total number of Back Tests required for training a robust systematic trading strategy consisting of allocating capital across a bag of stocks in order to minimise the estimated risk for a given expected return.The Trading StrategyWe are going to implement a standard Markowitz portfolio optimization strategy (MPT) that will have a number of tunable parameters. There are two quantities which MPT uses to make […]

  • How Active Learning can help you train your models with less Data
    by Charles Brecque on 09/10/2018 at 14:03

    (Source)Even with massive computational resources, training a Machine Learning model on large data sets can take hours, days and some times weeks which is expensive and a burden on your productivity. However, in most cases you do not need all the available data to train your models. In this article, we are going to compare data subsetting strategies and the impact they have on the performance of the models (training time and accuracy). We will implement them on the training of a SVM classifier […]

  • Early Stopping
    by Charles Brecque on 08/10/2018 at 12:48

    sourceSometimes it isn’t worth going to the end, especially in hyper-parameter tuningMost Machine Learning models have hyper-parameters which are fixed by the user in order to structure the training of these models on the underlying data sets. For example, you need to specify the depth and number of trees (among other hyper-parameters) when training a random forest. There are also many other “real world” examples of hyper-parameters. Once the hyper-parameters have been set, […]

  • Breaking into the Machine Learning Market
    by Charles Brecque on 08/10/2018 at 12:37

    sourceMachine Learning has been at the top of the Gartner Hype cycle since 2015 and the number of attendees at leading Machine Learning conferences such as NIPS has exploded over the past few years. This year’s conference sold out in just under 12 minutes!NIPS2018 The main conference sold out in 11 minutes 38 seconds — @NipsConferenceGartner Hype Curve 2015 (source)The hype around Machine Learning is fuelled by the Big tech companies who are fighting for […]

  • Hi Louis,
    by Charles Brecque on 04/10/2018 at 16:28

    Hi Louis,Thank you for the great article and links!I have a few comments/questions:It seems to me that most legacy organisations are trying to shift towards your AI first model because they already have the data and have decided to build in-house Data Science teams to leverage it.I also find that a lot of the legacy players are either weary of running a POC with an AI startup and when they do, it’s (slow) and mainly to learn about the space as opposed to actually buying the AI […]