Farruh’s Reflection on Week One Learning

Original article was published by K Farruh on Deep Learning on Medium


Farruh’s Reflection on Week One Learning

I started an AI course with unpack.AI, and fast.AI The Deep Learning for Coders is a fantastic course from fast.ai. The authors explained Deep Learning (DL) and Machine Learning (ML) in simple language.

In the first lesson, I have learned:

  1. What do I need to know a DL? It surprised me that the authors narrated that not need lots of math, lots of data.
  2. The power of DL. DL could be applied across many disciplines. These include the social and physical sciences, the arts, medicine, finance, scientific research, and many more.
  3. The advantage of DL in different areas, like:
  4. Natural language processing (NLP). When I am surfing an online shop and have any questions, the robot throws the chat by answering questions use NLP. Youtube has an automotive subtitle system that uses speech recognition. After the meeting, the secretary no needs to summarize or create a meeting recap. It can be done by DL and many other fields.
  5. Computer vision. Every day a dozen times, I am using face recognition to unlock my smartphone.
  6. Medicine. I hope one day, we can find a way to defeat any type of cancer with DL.
  7. Biology. The cell classifications task could be done by DL much better than any other tool.
  8. Image generation. In the future probably we will enjoy picture drown by AI.
  9. Recommendation systems. I am amazed by the Netflix recommendation system, every time it can offer me the right movie or serial.
  10. Playing games. How the AI can win humans in a Go game, impossible to believe.
  11. Robotics. I am looking forward to a time when robots could help me with housework and give my family more time.
  12. I can use the GPU and run code without a GPU device because of Google Colab.

Incredibly, I can run a model with high accuracy with several lines of code. With the fast.ai library, I can create image classification, text classification, prediction models.

Overall I enjoyed the first lesson, excited to take the next classes.