Learn Machine Learning - Workshop 3

This is a past event

94 people went

Location image of event venue

Details

***PLEASE REGISTER ON EVENTBRITE TO COME TO THIS MEET-UP***
https://www.eventbrite.co.uk/e/machine-learning-workshop-3-tickets-57539834193?aff=ebdssbdestsearch

***PLEASE BRING A LAPTOP****

***SPACES ARE VERY LIMITED ON THIS MEET-UP SO PLEASE CANCEL YOUR TICKET IF YOU CANNOT MAKE IT***

The April meet-up will have much more space so do not worry if you do not get ticket to attend this one.

*********************************************************************************
Schedule (Subject to some change):

6.30pm: Arrive

6.40pm: Welcome Talk.

6.45pm: Talk by Laurence Hubbard, head of data engineering at Compare the Market: 'Notes about Data Science from a Data Engineering perspective'.

7.00pm: Talk by Hayley Hubbard: 'How I Became a Data Scientist'.

7.15 -8.45pm: Break into groups and work through Competitions or tutorials. The suggested working options are below.

* Please note we currently do not have a food sponsor for this event so please bring your own. Sorry.

*********************************************************************************
Practical Work Options

1. Tutorials provided below.

2. Titanic Kaggle. (Beginner Level 1 Kaggle, Classification Problem).

3. House Price Kaggle. (Level 2 Kaggle).

4. Digit Recognizer. (Level 3 Kaggle, Computer Vision Fundamentals).

5. Facial Keypoints detection. (Level 4 Kaggle, Natural Language Processing).

6. Predicting Heart Disease challenge.

7. Data Science for Good: CareerVillage.org Kaggle.

*********************************************************************************
Who is this meetup for?

Ideally you have a background in software engineering, science or mathematics. All of these will make it easier to get stuck in solving the challenges.

*********************************************************************************
What is the aim of this meet-up?

The aim of these meetups is to provide an environment for you to teach yourself machine learning. The idea is to do this by working through tutorials or Kaggle competitions.

You will split into groups depending on what you want to work on. You can work as a team, individually or as a pair in these groups.

Any ability is welcome. This is a self-lead learning course and is based on everyone helping each other.

*********************************************************************************
Do I have to have been to the previous workshops to come along?

No. The groups will be split based on which challenges you would like to work on.

If you have not been to any of the other workshops you would get more out of the meetup by setting up your environments before coming along.

Please find the links of how to do soon the eventbrite link.

*********************************************************************************
How to work through these problems / Machine learning workflow:

1st Define the problem, your goal and what success would look like.

2nd Collect your data. You should split your data into training data and test data. Remove the label you are trying to predict from your test data and use this data to test the accuracy of your model. In the case of Kaggle this is done for you.

3rd Exploratory analysis and data prep. Visualise your train data set through bar charts etc to try to gain an understanding of what factors are important in predicting your label. Clean both data sets.

4th Predictive model logic. This is your machine learning model. Use any libraries out there to help you write an accurate model. Use this to predict the labels for your test data set.

5th Evaluate the accuracy of your model. If using Kaggle submit your result to the challenge and you will get your % accuracy.

6th Optimise and Improve. Re-iterate over steps 3 and 4 until your get your desired accuracy level.

*********************************************************************************
Suggested tutorials are found on the eventBrite.