Day 6 of Machine Learning in Review

Resources explored

For people whom need a little more structure I looked into some courses, here’s what I found so far:

• Practical Convolutional Neural Networks: The Course Overview | packtpub.com https://youtu.be/EgKQeuwmTEk

  • Looks amazingly useful for beginners
  • Love the examples
  • Authors seem well versed on the subject

• TensorFlow 1.x Deep Learning Recipes for Artificial Intelligence App: Course Overview | packtpub.com https://youtu.be/RMILgtLeNVc

  • Dives into Deep Learning and generally goes a step further that the previous course
  • The recipes for TensorFlow are my personal favourite

• Driven Data Competitions

  • their competition model almost ensures they always have a challenge open, currently 3–4 open
  • found this data source that might be useful for other purposes as well: https://data.worldbank.org/, although at first glance I wouldn’t know where to start if i was looking for a specific data set

• Kaggle Competitions

  • You can even see Suraj (author of #100DaysOfMLCode Challenge) competing on his channel: https://youtu.be/suRd3UzdBeo
    • watching this approach of Suraj would give you the insights necessary to his development process
  • I just had to pick a few really interesting challenges:
    • TrackML Particle Tracking Challenge I just love what CERN is doing, it’s great to see how the new tech can be useful for old research processes
    • Zillow’s Home Value Prediction This one I picked as would just love to be able to do this for myself.
    • iMaterialist Challenge (Furniture) This is a past challenge, but it makes for a great task that I could get involved in. Something like: Recognize valuable furniture in a second-hand/charity shop.

• Build a Neural Net in 4 Minutes

• Plant Identification - Picture This App

  • when it works it’s amazingly powerful

  • great idea of putting Neural Networks to work

  • the app still has some issues especially as I tried it out in the wild and was saying it had no connection, basically it needed loads of data to download

  • I love the use of community in this one in order to train the models

  • Some other resources on the subject:

    • Deep Learning for Plant Identification in Natural Environment
    • Plant identification based on very deep convolutional neural networks
    • Plant identification using deep neural networks via optimization of transfer learning parameters
    • DEEP-PLANT: PLANT IDENTIFICATION WITH CONVOLUTIONAL NEURAL NETWORKS
    • Original Article: This App Will Help You Identify the Plants and Animals You See in Nature

• AI Experiments

  • One such experiment: A.I. Experiments: Visualizing High-Dimensional Space