DebKR

To the Stars

Blog

Diving Into Data: Personalised Training Plan (work in progress)

Diving Into Data: Personalised Training Plan (work in progress)

By

training-planBelow is my personalised training plan to move into a career in Data/Analytics with Machine Learning, with specific reference to consumer-direct/e-commerce business. To be implemented throughout March- November 2016.

Areas to cover

Mathematics: Linear algebra, Multivariate calculus

Probability and Statistics: Inferential stats, Descriptive stats, Regression analysis (linear, logistic), Cluster analysis, Classification and regression tree model (CART), Hypothesis testing, Decision trees, Predictive modelling/forecasting [Read more…] about Diving Into Data: Personalised Training Plan (work in progress)

Resources: Programming / Data Science / Analytics (work in progress)

Resources: Programming / Data Science / Analytics (work in progress)

By

Universities
Stanford University: http://cs.stanford.edu/research/computer-systems
Review Piazza for participating courses/universities & professors in relevant fields: Computer Science – Mathematics – Engineering – Economics
Wharton Customer Analytics Initiative: http://wcai.wharton.upenn.edu/

Companies
Marketing Platform http://rocketfuel.com/
Business strategy through data http://www.meltwater.com
Data management software http://www.sas.com
Data science & consulting http://www.persontyle.com/

Networking
Deep Learning London meetup http://www.meetup.com/Deep-Learning-London/
[Read more…] about Resources: Programming / Data Science / Analytics (work in progress)

Resources: Artificial Intelligence / Machine Learning (work in progress)

Resources: Artificial Intelligence / Machine Learning (work in progress)

By

AIUniversities
Stanford Computer Science (AI Research): http://cs.stanford.edu/research/ai
Stanford AI Lab: http://ai.stanford.edu/
Stanford Search: http://www.stanford.edu/search/
Stanford AI100: One Hundred Year Study on Artificial Intelligence

Articles
Why big tech companies are open-sourcing their AI systems
How many computers does it take to identify a cat?

People
Andrew Ng http://www.andrewng.org/
Yaser S. Abu-Mostafa http://work.caltech.edu/
See also my post here (various sources as given): Most Influential Data Scientists (Two Lists) (includes Machine Learning)
Dr. Francois M. Vanderseypen: https://github.com/Orbifold/ \\ http://www.orbifold.net/ \\

Software
TensorFlow https://www.tensorflow.org/ (Google says: “we believe [TensorFlow] to be one of the best machine learning toolboxes in the world, we hope to create an open standard for exchanging research ideas and putting machine learning in products. Google engineers really do use TensorFlow in user-facing products and services, and our research group intends to share TensorFlow implementations along side many of our research publications.”) (includes ‘Get Started’ and ‘Tutorials’ sections.)
Orbifold ML algos on GitHub: https://github.com/Orbifold/XAct

Learning: Recommended Courses
Article: How do I learn machine learning?
Article: 7 Free Machine Learning Courses

Glossaries
Machine Learning – Glossary of Terms, Stanford Artificial Intelligence Lab
Machine Learning Dictionary, School of Computer Science and Engineering, University of New South Wales

 

Where to Start in Programming 2

Where to Start in Programming 2

By

programming-pythonOption 1: the intention is to create something specific right now (start with the end in mind)
  • Web developing: HTML, CSS, JavaScript, PHP
  • App developing: Java for Android, or Objective C for iOS
  • Mainstream programming, various uses: C, C#, Java, Perl, Python, Ruby

Option 2: the intention is to become a programmer, we can worry about what to create later (eat, sleep, code)

  • To learn to code efficiently, at machine-level, choose C
    • Advantages: you learn the fundamentals of coding; see how the language interacts with the hardware; requires discipline so teaches you to code efficiently
    • Disadvantage: strict; not really beginner-friendly; longer learning time than some other languages; can be frustrating (e.g. debugging for errant semicolons)
  • To learn the principles of Object Oriented Programming (OOP), choose Java
    • Advantages: highly practical language; learning Java means you’re learning the OOP principles – allows for easier/faster learning of more modern languages like C++, Perl, Python, PHP; taught for free via Stanford Uni’s open-access Programming Methodology course; used extensively across many platforms; used in Android app development; long history of use means lots of standard examples available
    • Disadvantages: ??

[Read more…] about Where to Start in Programming 2

Where to Start in Programming

Where to Start in Programming

By

programmingIn school we were taught Basic. As a maths undergrad I studied programming, primarily C++. So long ago now, I don’t remember much at all.

If I were to learn programming now, it’s a minefield: What language to concentrate on? What’s the best route for a beginner to follow? What hardware requirements are there so I can get going (economically)? Are there different routes one could follow that determine what kind of programming they could end up doing? Or is there a generic language I can start with, then branch out from there? How long will it take me to learn (to reasonable level of competency) any given programming language?

Some general thoughts and research (in no particular order):

  • Finance systems implementation (current project): Unit4 (formerly CODA) Dream accounting software is “built on Windows technology”
  • Windows OS mainly uses C, C++ with some C# for Windows, also some parts in assembly language
  • XML (Extensible Markup Language): A metalanguage, powerful option for importing/exporting data between different users/applications. Through a set of pre-defined rules, documents or data can be formatted which is machine-readable but also remains readable by humans. Referred to in the book Digital Accounting: The Effects of the Internet and ERP on Accounting by Ashutosh Deshmukh. Used by various governments in ‘tagged’ accounting submissions. Not widely deployed by small businesses despite it’s relative ease to implement. Could be used for (dare I say it) passing information between departments

[Read more…] about Where to Start in Programming