Difference between revisions of "ML curriculum"
From Fiamma
Jump to navigationJump to search| Line 3: | Line 3: | ||
* [https://engineering.ucsb.edu/~shell/che210d/python.pdf Short intro to Python in 60 pages], 50% done. | * [https://engineering.ucsb.edu/~shell/che210d/python.pdf Short intro to Python in 60 pages], 50% done. | ||
* [http://neuralnetworksanddeeplearning.com/chap1.html Neural networks and deep learning book] -- probably the one I'll be using | * [http://neuralnetworksanddeeplearning.com/chap1.html Neural networks and deep learning book] -- probably the one I'll be using | ||
| + | * [https://github.com/ZuzooVn/machine-learning-for-software-engineers/blob/master/README.md ML for Software Engineers] also looks wonderful | ||
<hr> | <hr> | ||
* [https://pandas.pydata.org/pandas-docs/stable/10min.html 10 minutes to pandas] | * [https://pandas.pydata.org/pandas-docs/stable/10min.html 10 minutes to pandas] | ||
| Line 8: | Line 9: | ||
* [https://github.com/p-i-/machinelearning-IRC-freenode/blob/master/Resources/ArticlesIntroductory.md ArticlesIntroductiory] ##machinelearning's intro articles, esp math. | * [https://github.com/p-i-/machinelearning-IRC-freenode/blob/master/Resources/ArticlesIntroductory.md ArticlesIntroductiory] ##machinelearning's intro articles, esp math. | ||
* http://www.deeplearningbook.org/ nice intro book | * http://www.deeplearningbook.org/ nice intro book | ||
| − | |||
== More interesting stuff == | == More interesting stuff == | ||
* [http://www.holehouse.org/mlclass/ Andrew Ng's Stanford ML course, unofficial notes] as Ersatz for the Coursera video-course. | * [http://www.holehouse.org/mlclass/ Andrew Ng's Stanford ML course, unofficial notes] as Ersatz for the Coursera video-course. | ||
Revision as of 14:27, 16 May 2018
Using this as base:
Basics
- Short intro to Python in 60 pages, 50% done.
- Neural networks and deep learning book -- probably the one I'll be using
- ML for Software Engineers also looks wonderful
- 10 minutes to pandas
- Scipy lecture notes -- looks excellent
- ArticlesIntroductiory ##machinelearning's intro articles, esp math.
- http://www.deeplearningbook.org/ nice intro book
More interesting stuff
- Andrew Ng's Stanford ML course, unofficial notes as Ersatz for the Coursera video-course.
- Later think about this, but I'm not sure I like those links. TODO
- https://github.com/p-i-/machinelearning-IRC-freenode/blob/master/Resources/GettingStarted.md Coursera articles, again from #machinelearning
- https://developers.google.com/machine-learning/crash-course/ Google ML crash course
- https://github.com/p-i-/machinelearning-IRC-freenode/blob/master/Resources/Main.md Main ##ml "Main"
- https://ai.google/education google AI
- Awesome machine learning
- https://github.com/floodsung/Deep-Learning-Papers-Reading-Roadmap/blob/master/README.md
- http://course.fast.ai/ nice video AI course
Мария, [05.03.18 21:55] - TensorFlow https://www.tensorflow.org/ -- Шарить что такое граф -- Шарить как работать с сессией (как создавать, как запускать операции на выполнение) -- Понимать вообще глобальную архитектуру приложения на TensorFlow -- Хоть чуть-чуть иметь представление как делать backprop через граф - Как дебажить обучение -- Понять из графика лосса что происходит с обучением -- Overfitting, underfitting, vanishing/exploding gradient - Алгоритмы оптимизации -- SGD, SGD with momentum, RMSProp, Adam - Современные архитектуры -- VGG, Segnet, ResNet, Inception, Inception-ResNet, NasNet -- Понимать зачем нужна каждая из них и для чего ее создавали - OpenCV -- Гистограмное выравнивание -- Cascade Classifier - dlib -- Просто знать что там можно искать лица и ключевые точки лица - Python -- Библиотеки numpy, matplotlib -- Уметь грузить/сохранять файлы Мария, [05.03.18 21:55] ++++++++курс на курсере конечно
