Natural language processing open seminar For Tensorflow usage
- 3. We Studied lots of Algorithms in machine Learning, in particular with Deep Learning!
- 4. We Studied lots of Algorithms in machine Learning, in particular with Deep Learning! There are many libraries for those algorithms
- 5. There are many libraries and environment for implementing machine learning and deep learning and Tensorflow is popular and steadily develops.
- 7. There are two types of mode in computational operation.
- 8. There are two types of mode in computational operation. Define and Run Define by Run
- 9. Tensorflow consists of Node and Edge.
- 10. First all, Let’s see example about define and run.
- 11. Suppose Tensorflow implements + operation(2 + 3 = 5). You can evaluate your model of the Computational Graph by tf.Session. The computational graph is evaluated by calling run() method of tf.Session.
- 12. Suppose Tensorflow implements add and multiply operation(1+2 and 1*2).
- 13. Suppose Tensorflow implements add and multiply operation one more time.
- 14. Suppose Tensorflow implements add and multiply operation one more time.
- 15. Suppose Tensorflow implements add and multiply operation one more time.
- 16. Suppose Tensorflow implements add and multiply operation one more time.
- 17. Suppose Tensorflow implements add and multiply operation one more time.
- 18. Tensorflow evaluate operation efficiently on execution time.
- 19. Tensorflow evaluate operation efficiently on execution time.
- 20. Eager execution is comfortable with programming(define by run)
- 21. Eager execution is comfortable with programming(define by run)
- 22. Eager execution has a benefit that all the functionality of the host language is available while your models is executing as follows.
- 23. During eager execution, you can use tf.GradientTape to trace operations for computing gradient later.
- 24. Traditional Machine Learning and Deep Learning have Learning part. But, here we automatically generate data and learn the regression model both normal one and neural network.
- 25. what kind of cost function is when you learn regression model?
- 26. How to learn the model on what kind of update function.
- 27. What is the linear regression?
- 28. The stage of Data generation
- 29. Construction of model
- 30. Construction of model
- 31. Construction of model
- 32. Start of learning
- 33. The stage of learning
- 34. What about sin(x) function for prediction with Neural Network?
- 35. The stage of data generation
- 43. (Bonus)Let’s evaluate regression model on eager execution!
- 44. (Bonus)Let’s evaluate regression model on eager execution!
- 45. (Bonus)Let’s evaluate regression model on eager execution!
- 46. (Bonus)Let’s evaluate regression model on eager execution!