Recommended
Recommended
More Related Content
What's hot
What's hot (20)
Similar to Artificial Intelligence, Machine Learning and Deep Learning with CNN
Similar to Artificial Intelligence, Machine Learning and Deep Learning with CNN (20)
Recently uploaded
Recently uploaded (20)
Artificial Intelligence, Machine Learning and Deep Learning with CNN
- 1. Artificial Intelligence, Machine Learning,Deep Learning Mojammel Haque Mithu AI Researcher & Developer @ SemanticsLab mojammelbrur@gmail.com May 06, 2019
- 2. Outline What is AI? What is Machine Learning? Types of Machine Learning Machine Learning Algorithms What is Deep learning? Machine Learning vs Deep learning Basic neural network Forward pass and Backward pass Convolution Neural Network(CNN) with example
- 3. What is Artificial intelligence (AI)? Artificial intelligence (AI) is the simulation of human intelligence processes by machines, especially computer systems. These processes include learning (the acquisition of information and rules for using the information), reasoning (using rules to reach approximate or definite conclusions) and self-correction. Particular applications of AI include expert systems, speech recognition and machine vision.
- 4. Example of Artificial intelligence (AI)
- 5. What is Machine Learning? Machine Learning is a field of computer science that gives computer systems the ability to “learn” (i.e progressively improve performance on specific task) with data, without being explicitly programmed.-WIKI
- 6. Example of Machine Learning Mars rover
- 7. Types of Machine Learning
- 8. Machine Learning Algorithms 1. Linear Regression 2. Logistic Regression 3. Linear Discriminant Analysis 4. Classification and Regression Trees 5. Naive Bayes 6. K-Nearest Neighbors 7. Learning Vector Quantization 8. Support Vector Machines 9. Bagging and Random Forest 10. Boosting and AdaBoost
- 9. Support Vector Machine Support Vector Machine is supervised machine learning algorithm which is mostly used in classification problems. we perform classification by finding the hyper-plane that differentiate the two classes very well.
- 10. Linear Regression Linear Regression is a machine learning algorithm based on supervised learning. It performs a regression task
- 11. K-Nearest Neighbors The KNN algorithm assumes that similar things exist in close proximity. In other words, similar things are near to each other.
- 12. What is Deep learning? Deep Learning is a subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks.
- 13. What Deep Learning can do? 1. It can also prescribe medicine used in medication. 2. Computer vision and pattern recognition 3. Robotics — Deep Learning systems have been taught to play games and even made to taught WIN games. 4. Facial recognition 5. Precision agriculture 6. Fashion technology 7. Autonomous vehicles 8. Drone and 3D mapping 9. Post estimation in Sports analytics & Retail markets 10.Security & Surveillance
- 14. Machine Learning Vs Deep Learning The key difference between deep learning vs machine learning stems from the way data is presented to the system. Machine learning algorithms almost always require structured data, whereas deep learning networks rely on layers of the ANN (artificial neural networks).
- 15. Machine Learning Vs Deep Learning
- 16. Structure of a simple neural network
- 17. Feed-Forward Networks • Predictions are fed forward through the network to classify 17
- 18. Feed-Forward Networks • Predictions are fed forward through the network to classify 18
- 19. Feed-Forward Networks • Predictions are fed forward through the network to classify 19
- 20. Feed-Forward Networks • Predictions are fed forward through the network to classify 20
- 21. Feed-Forward Networks • Predictions are fed forward through the network to classify 21
- 22. Feed-Forward Networks • Predictions are fed forward through the network to classify 22
- 23. Error Backpropagation • We will do gradient descent on the whole network. • Training will proceed from the last layer to the first. 23
- 24. Error Backpropagation • Introduce variables over the neural network 24
- 25. Error Backpropagation • Introduce variables over the neural network – Distinguish the input and output of each node 25
- 27. Error Backpropagation 27 Training: Take the gradient of the last component and iterate backwards
- 28. Error Backpropagation 28 Empirical Risk Function
- 29. Error Backpropagation 29 Optimize last layer weights wkl Calculus chain rule
- 30. Error Backpropagation 30 Optimize last layer weights wkl Calculus chain rule
- 31. Error Backpropagation 31 Optimize last layer weights wkl Calculus chain rule
- 32. Error Backpropagation 32 Optimize last layer weights wkl Calculus chain rule
- 33. Error Backpropagation 33 Optimize last layer weights wkl Calculus chain rule
- 34. Error Backpropagation 34 Optimize last hidden weights wjk
- 35. Error Backpropagation 35 Optimize last hidden weights wjk Multivariate chain rule
- 36. Error Backpropagation 36 Optimize last hidden weights wjk Multivariate chain rule
- 37. Error Backpropagation 37 Optimize last hidden weights wjk Multivariate chain rule
- 38. Error Backpropagation 38 Optimize last hidden weights wjk Multivariate chain rule
- 39. Error Backpropagation 39 Repeat for all previous layers
- 40. Error Backpropagation 40 Now that we have well defined gradients for each parameter, update using Gradient Descent
- 41. Error Back-propagation • Error backprop unravels the multivariate chain rule and solves the gradient for each partial component separately. • The target values for each layer come from the next layer. • This feeds the errors back along the network. 41
- 42. Convolutional Neural Network (CNN)
- 79. Thank you
Editor's Notes
- \vec{\theta}=\{w_{ij}, w_{jk}, w_{kl}\}
- \vec{\theta}=\{w_{ij}, w_{jk}, w_{kl}\}
- \vec{\theta}=\{w_{ij}, w_{jk}, w_{kl}\}
- \vec{\theta}=\{w_{ij}, w_{jk}, w_{kl}\}
- \frac{\partial R}{\partial w_{kl}}=\frac{1}{N}\sum_n\left[\frac{\partial L_n}{\partial a_{l,n}}\right]\left[\frac{\partial a_{l,n}}{\partial w_{kl}}\right]
- \frac{\partial R}{\partial w_{kl}}=\frac{1}{N}\sum_n\left[\frac{\partial \frac{1}{2}(y_n-g(a_{l,n}))^2}{\partial a_{l,n}}\right]\left[\frac{\partial z_{k,n}w_{kl}}{\partial w_{kl}}\right]&=&\sum_n\left[-(y_n-z_{l,n})g'(a_{l,n})\right]z_{k,n}\\&=&\sum_n\delta_nz_{k,n}
- \frac{\partial R}{\partial w_{kl}}=\frac{1}{N}\sum_n\left[\frac{\partial \frac{1}{2}(y_n-g(a_{l,n}))^2}{\partial a_{l,n}}\right]\left[\frac{\partial z_{k,n}w_{kl}}{\partial w_{kl}}\right]&=&\sum_n\left[-(y_n-z_{l,n})g'(a_{l,n})\right]z_{k,n}\\&=&\sum_n\delta_nz_{k,n}
- \frac{\partial R}{\partial w_{kl}}=\frac{1}{N}\sum_n\left[\frac{\partial \frac{1}{2}(y_n-g(a_{l,n}))^2}{\partial a_{l,n}}\right]\left[\frac{\partial z_{k,n}w_{kl}}{\partial w_{kl}}\right]&=&\sum_n\left[-(y_n-z_{l,n})g'(a_{l,n})\right]z_{k,n}\\&=&\sum_n\delta_nz_{k,n}
- \frac{\partial R}{\partial w_{kl}}=\frac{1}{N}\sum_n\left[\frac{\partial \frac{1}{2}(y_n-g(a_{l,n}))^2}{\partial a_{l,n}}\right]\left[\frac{\partial z_{k,n}w_{kl}}{\partial w_{kl}}\right]&=&\sum_n\left[-(y_n-z_{l,n})g'(a_{l,n})\right]z_{k,n}\\&=&\sum_n\delta_nz_{k,n}
- \frac{\partial R}{\partial w_{jk}}=\frac{1}{N}\sum_n\left[\sum_l\delta_l w_{kl}g'(a_{k,n})\right]\left[z_{j,n}\right]\\
- \frac{\partial R}{\partial w_{ij}}&=&\frac{1}{N}\sum_n\left[\frac{\partial L_n}{\partial a_{j,n}}\right]\left[\frac{\partial a_{j,n}}{\partial w_{ij}}\right]=\frac{1}{N}\sum_n\left[ \sum_k\delta_{k,n}w_{jk}g'(a_{j,n}) \right]z_{i,n} =\frac{1}{N} \sum_n\delta_{j,n}z_{i,n}
- \frac{\partial R}{\partial w_{jk}}=\frac{1}{N}\sum_n\left[\sum_l\delta_l w_{kl}g'(a_{k,n})\right]\left[z_{j,n}\right]\\