|Learn From Top Kagglers|
|Written by Sue Gee|
|Tuesday, 10 April 2018|
A course in which top Kaggler's share their expertise has just restarted on Coursera. It is part of a new set of courses that comprise a recently introduced Advanced Machine Learning Specialization.
The title of this course, How to Win a Data Science Competition: Learn from Top Kagglers tells you most of what you need to know before you sign up. Like the other six courses that make up the specialization it comes from the National Research University Higher School of Economics (HSE), one of the top research universities in Russia. The five instructors for this course are: Dmitry Ulyanov, Alexander Guschin, Mikhail Trofimov, Dmitry Altukhov and Marios Michailidis, all of them prominent members of the Kaggle community.
In case you need reminding, Kaggle, which joined Google in 2017, was formed in 2010 as a platform for predictive modelling and analytics competitions. It recently launched int own educational platform, see my report Introducing Kaggle Learn for those wanting to get into machine learning, R programming, data visualisation and deep learning.
This course, in the other hand, is primarily for those who want to join in Kaggle contests. The course description states:
If you want to break into competitive data science, then this course is for you! Participating in predictive modelling competitions can help you gain practical experience, improve and harness your data modelling skills in various domains such as credit, insurance, marketing, natural language processing, sales’ forecasting and computer vision to name a few. At the same time you get to do it in a competitive context against thousands of participants where each one tries to build the most predictive algorithm. Pushing each other to the limit can result in better performance and smaller prediction errors. Being able to achieve high ranks consistently can help you accelerate your career in data science.
Students can expect to spend 6-10 hours per week over 6 weeks culminating in a final project which is itself a Kaggle contest in which you compete against other course participants. The disclaimer for this class states:
This is not a machine learning course in the general sense. This course will teach you how to get high-rank solutions against thousands of competitors with focus on practical usage of machine learning methods rather than the theoretical underpinnings behind them.
Given the highly specific goal of the class many of its students will probably treat it as a standalone MOOC. For others however it will come as the second component of Coursera's Advanced Machine Learning Specialization which is designed to provide an introduction to deep learning, reinforcement learning, natural language understanding, computer vision and Bayesian methods for students who are already familiar with machine learning fundamentals. The program description states:
You will master your skills by solving a wide variety of real-world problems like image captioning and automatic game playing throughout the course projects. You will gain the hands-on experience of applying advanced machine learning techniques that provide the foundation to the current state-of-the art in AI.
The starting point for the specialization is Introduction to Deep Learning (6 weeks, 6-10 hrs per week) and its prerequisites are basic knowledge of Python; basic linear algebra and probability and basic knowledge of machine learning including linear regression, logistic regression, gradient descent for linear models, the problem of overfitting and regularization for linear models.
The remaining components of the Specialization, which can be taken in any order and also be taken as standalone courses are:
Bayesian Methods for Machine Learning 6 weeks, 6 hrs per week
Natural Language Processing 5 weeks, 4-5 hrs per week
Practical Reinforcement Learning 6 weeks, 3-9 hrs per week
Deep Learning in Computer Vision 5 weeks Starts April 2018
Addressing Large Hadron Collider Challenges by Machine Learning 5 weeks Starts May 2018
The first five courses in the Advanced Machine Learning have April 9th as the most recent start date and will re-run on a regular basis. As this isn't the first presentation of How to Win a Data Science Competition we have the benefit of student ratings and reviews. Overall it has 4.7 stars from 168 ratings and 87% awarded it a full 5-star rating. Here's a helpful 5-star review:
Just like the course 1 of this specialization, the pace of course 2 on practical data science competition is very fast, therefore the quizzes and assignments are indeed necessary and very helpful (even for the final project). Some of the programming tasks & final project are quite hard and time consuming. But it is worthwhile to grasp the practical knowledge/skills and work on the real-life programming solution.
One aspect that lost the course stars was the quality of the instructors voices and their use of English as summed up in this 2-star review:
Content is really good. But delivery is at times incomprehensible. Assignments questions are also not very clear.
A different criticism was voiced in this 1-star review:
Should have been labeled "How to Cheat a Data Science competition". An entire week is dedicated to Data Leakage and how to exploit it rather than in the spirit of the competition how to create a model that actually solves the problem.
To take part fully in this course you will need to pay - the audit route only gives access to the video lectures and some of the assignments. There is 7-day free trial beyond which you pay $49 per month - so the faster you work the less you pay and the fee covers all the courses. Financial aid is available but applications take at least 15 days to be reviewed.
or email your comment to: email@example.com
|Last Updated ( Tuesday, 10 April 2018 )|