I have over 18 years experience in the software industry and for a year or more now I’ve been trying to learn AI/ML but I’ve not really found the right training materials. I was thinking of doing a MS in machine learning. Anyone done something similar before? Is it worth it? Money is not a concern. It’s more about learning a new skill from scratch.
Coursera
Take Andrew Ng’s courses in ML
Ok I will check them out
Don't bother going to school for this - it's a waste of time and money. Why are you not able to find the right training material? There is no shortage of great material on the web.
I am in the same boat. Can someone point to the right material or course they found to get in into Al/ML field
I considered the same but I decided against it because of the reasons stated above. I learnt mostly through books and online courses. A few resources that I found to be helpful while learning; Deep Learning (Adaptive Computation and Machine Learning series) https://a.co/d/4LysL1s Inside Deep Learning: Math, Algorithms, Models https://a.co/d/fjg9OMJ Mathematics for Machine Learning https://a.co/d/aidth19 https://www.coursera.org/specializations/machine-learning-introduction https://www.deeplearning.ai/courses/deep-learning-specialization/ https://www.coursera.org/specializations/mathematics-machine-learning https://youtube.com/playlist?list=PLAqhIrjkxbuWI23v9cThsA9GvCAUhRvKZ&si=uQeXYLTGLE5UEbac
Excellent. Thank you.
I had CHATGPT make me a 12 weeks learning plan with references to links, it’s legit good.
please share here if you don't mind.
Sure, creating a learning plan for deep learning involves several key steps: 1. **Foundational Knowledge:** - Understand basic concepts of linear algebra, calculus, and probability. - Familiarize yourself with programming languages like Python. 2. **Core Concepts of Machine Learning:** - Learn about supervised and unsupervised learning. - Understand key machine learning algorithms and their applications. 3. **Introduction to Neural Networks:** - Study basic concepts of neural networks. - Learn about activation functions, weights, and biases. 4. **Deep Learning Frameworks:** - Choose a deep learning framework (e.g., TensorFlow, PyTorch) and become proficient in it. - Practice building simple neural networks. 5. **Convolutional Neural Networks (CNNs):** - Understand CNNs and their applications in image processing. - Implement CNNs on image datasets. 6. **Recurrent Neural Networks (RNNs):** - Learn about RNNs for sequential data. - Implement RNNs on text or time-series data. 7. **Transfer Learning:** - Explore transfer learning techniques and pre-trained models. - Implement transfer learning on your own projects. 8. **Generative Adversarial Networks (GANs):** - Study GANs for generative tasks. - Implement GANs for image generation. 9. **Natural Language Processing (NLP):** - Dive into NLP concepts and applications. - Work with language models and sentiment analysis. 10. **Deploying Models:** - Understand the basics of model deployment. - Deploy a deep learning model using frameworks like Flask or Docker. 11. **Continuous Learning:** - Stay updated on the latest research in deep learning. - Participate in online communities, forums, and conferences. 12. **Projects:** - Apply your knowledge to real-world projects. - Build a portfolio showcasing your deep learning projects. Remember to practice regularly, work on projects, and seek out additional resources such as online courses, tutorials, and research papers to deepen your understanding. Adjust the pace based on your learning style and goals.
Coursera
It could potentially help you get your foot in the door at some companies, but I’d imagine a MS in ML is going to cover a lot of outdated content considering how quickly things are moving. For example, most of the gen AI we see right now uses transformers which Google invented in 2017. I’m sure the fundamentals will still be valuable but you could probably learn those along with more relevant skills by doing something other than a MS. If you just love learning or want to have the MS then sure, do it. But there are probably more optimal ways to reach your goal.
This ^ the course material will most likely be outdated. It’s one of the annoying things about some advanced CS courses - the content is not really relevant anymore
Yes it’s the fundamentals that have been hard to learn on any courses I found.