2018 Global Azure Bootcamp Azure Machine Learning for neural networks
Download as PPTX, PDF2 likes143 views
The document outlines a comprehensive hands-on lab agenda focused on Azure Machine Learning and neural networks, emphasizing their applications in machine learning. It explores fundamental concepts, algorithms, and processes associated with machine learning, including regression, clustering, and reinforcement learning. Additionally, the document provides practical examples and references for implementing these techniques using Azure Machine Learning Studio.
2018 Global Azure Bootcamp Azure Machine Learning for neural networks
- 1. Azure Machine Learning – Neural Networks Setu Chokshi, AI MVP
- 2. Azure Machine Learning – Hands on Lab
- 4. Why? Discover reason behind success, failure Understand customers, products Plan future Experiment meaningfully Improve performance Run on analytics
- 5. Data science
- 6. Machine learning ≣ data mining
- 7. Examples of Machine Learning
- 8. How does machine learning help? There are only 5 questions that machine learning can help answer Source: Data Science For Beginners - 5 Questions Data Science Answers by Brandon Rohrer
- 9. 1. Is this A or B? Real-Time Human Pose Recognition in Parts from a Single Depth Image
- 10. 2. Is this Weird? Is this Weid? Anomaly detection algorithms
- 11. 3. How much? How many? How many? How much? Regression algorithms
- 12. 4. How is this organized? How is this organized? Clustering algorithms
- 13. 5. What should I do now? What should I do now? Reinforcement learning algorithms
- 14. Machine learning process and algorithms
- 15. How does it work Algorithm Your data Computer Your answer Recipe Ingredients Blender Smoothie
- 16. 1. Define & initialise a model 2. Train model (process cases) 3. Validate model …by scoring (making predictions) a test data set and evaluating the results 4. Use it: Explore or Deploy …visualise and study …deploy as a (web) service 5. Update and revalidate How?
- 19. Lets do a simple linear regression
- 20. Statistics
- 22. Datasets
- 24. Lets open Azure Machine Learning Studio
- 25. Perceptron X1 X2 X3 W11 W21 W31 W = Weight is the strength of the connection between nodes b Summation = w11 * X1 + w21 * X2 + w31 * X3 + b Output = 0 if summation <= Threshold = 1 if summation > Threshold OutputAct
- 26. Activation functions Output = 0 if summation <= Threshold = 1 if summation > Threshold 0 1 -ve +ve Step Function
- 27. Activation Functions Output = c* Input -ve +ve Linear Function
- 28. Activation functions Output = Range between (0, 1) 0 1 -ve +ve 0.5 Sigmoid Function 𝐴𝑐𝑡𝑖𝑣𝑎𝑡𝑖𝑜𝑛 = 1 1 + 𝑒−𝑥
- 29. Activation functions Output = max(0,input) 0 1 -ve +ve 0.5 Rectified Linear Unit (ReLU)
- 30. Perceptron X1 X2 X3 W = Weight is the strength of the connection between nodes Summation = w11 * X1 + w21 * X2 + w31 * X3 + b Output = 0 if summation <= Threshold = 1 if summation > Threshold Output Hidden h1 1 W11 W21 W31 Act b h1 = w11 * X1 + w21 * X2 + w31 * X3 + 1 Output = 0 if summation <= Threshold = 1 if summation > Threshold
- 31. One layer artificial neural network Y1 Y2 X1 X2 X3 h1 h2 h3 h4 Information Transfer Input Values Calculator – Activations Output (Activation) 1 2 3 4 Input Hidden Output h1
- 32. Now lets try the hands on
- 33. Handwriting Recognition with ANN
- 35. What is a convolution
- 36. Now onto the Azure Machine Learning
- 37. What does it look like? Reference: Gradient-based learning applied to document recognition , Proc. IEEE 86(11): 2278–2324, 1998. LENETArchitecture
- 38. A bit of history…Alexnet (2012)
- 39. A bit of history…VGG (2014)
- 40. A bit of history…Inception (aka Googlenet, 2014)
- 41. What is the network learning? Reference: Visualizing and Understanding Convolutional Networks, 2014
- 42. What is the network learning?
- 44. What is custom vision?
- 45. Hands on with Azure Machine Learning
- 46. Lets start with handwriting
- 47. ..and now lets do cats and dogs
- 48. More to come on Azure AI Offerings
Editor's Notes
- #29: between X values -2 to 2, Y values are very steep. Which means, any small changes in the values of X in that region will cause values of Y to change significantly. So this function has a tendency to bring the Y values to either end of the curve. If you notice, towards either end of the sigmoid function, the Y values tend to respond very less to changes in X. What does that mean? The gradient at that region is going to be small. It gives rise to a problem of “vanishing gradients”. Hmm. So what happens when the activations reach near the “near-horizontal” part of the curve on either sides? Gradient is small or has vanished ( cannot make significant change because of the extremely small value ). The network refuses to learn further or is drastically slow ( depending on use case and until gradient /computation gets hit by floating point value limits ).
- #37: https://pastebin.com/JRbz24JX https://pastebin.com/kRmHT4JW