Machine Learning Algorithms | Machine Learning Tutorial | Data Science Algorithms | Simplilearn
45 likes•13,530 views
The document provides an overview of machine learning, detailing its definition, processes, and various algorithms including linear regression, logistic regression, decision trees, random forests, and k-nearest neighbors. It explains real-world applications of machine learning in fields like healthcare, finance, and marketing. Additionally, the document illustrates implementation steps for several algorithms, including data preprocessing, model training, and evaluation metrics.
Machine Learning Algorithms | Machine Learning Tutorial | Data Science Algorithms | Simplilearn
- 2. Machine Learning Example How does Snapchat filter work?
- 4. Machine Learning Example Matrix of numbers
- 5. Machine Learning Example Matrix of numbers Looks for color patterns on face
- 6. Machine Learning Example Matrix of numbers Looks for color patterns on face Points align themselves and look for areas of contrast
- 7. Machine Learning Example Matrix of numbers Looks for color patterns on face Points align themselves and look for areas of contrast
- 8. What’s in it for you? Types of Machine Learning Algorithms Real World Applications of Machine Learning Processes involved in Machine Learning What is Machine Learning? Logistic Regression Linear Regression Decision Tree and Random forest K Nearest Neighbors Popular Algorithms with Hands On Demo
- 9. Real World Applications of Machine Learning
- 10. Real World Applications of Machine Learning Face Recognition Healthcare Industry Weather Forecasting Produce a Web Series Prepare a new Drink Siri and Cortana
- 11. What is Machine Learning?
- 12. What is Machine Learning? Machine Learning is the science of making computers learn and act like humans by feeding data and information without being explicitly programmed.
- 13. What is Machine Learning? Past Data Machine Learning is the science of making computers learn and act like humans by feeding data and information without being explicitly programmed.
- 14. What is Machine Learning? Past Data Analyses Machine Learning is the science of making computers learn and act like humans by feeding data and information without being explicitly programmed.
- 15. What is Machine Learning? Past Data Analyses Trains Machine Learning is the science of making computers learn and act like humans by feeding data and information without being explicitly programmed.
- 16. What is Machine Learning? Past Data Analyses Predicts Trains Machine Learning is the science of making computers learn and act like humans by feeding data and information without being explicitly programmed. Output
- 17. Processes involved in Machine Learning Data Gathering 7Machine Learning Processes
- 18. Processes involved in Machine Learning Data Gathering 7Machine Learning Processes Data Pre-Processing
- 19. Processes involved in Machine Learning Data Gathering 7Machine Learning Processes Data Pre-Processing Choose Model
- 20. Processes involved in Machine Learning Data Gathering 7Machine Learning Processes Data Pre-Processing Choose Model Train Model
- 21. Processes involved in Machine Learning Data Gathering 7Machine Learning Processes Data Pre-Processing Choose Model Train Model Test Model
- 22. Processes involved in Machine Learning Data Gathering 7Machine Learning Processes Data Pre-Processing Choose Model Train Model Test Model Tune Model
- 23. Processes involved in Machine Learning Data Gathering 7Machine Learning Processes Data Pre-Processing Choose Model Train Model Test Model Tune Model Prediction
- 24. Types of Machine Learning Algorithms VSupervised Learning Un-Supervised Learning Reinforcement Learning Machine Learning Association • Market Basket Analysis • Text Mining • Face Recognition Classification • Fraud Detection • Email Spam Detection • Image Classification Regression • Weather Forecasting • Risk Assessment • Score Prediction Clustering • Medical Research • City Planning • Targeted Marketing Reinforcement Learning • Gaming • Robot Navigation • Stock Trading • Assembly Line Processes
- 25. Popular Algorithms in Machine Learning Algorithm 1 Linear Regression
- 26. Popular Algorithms in Machine Learning Algorithm 1 Algorithm 2 Linear Regression Logistic Regression
- 27. Popular Algorithms in Machine Learning Algorithm 1 Decision Tree Algorithm 2 Algorithm 3 Linear Regression Logistic Regression
- 28. Popular Algorithms in Machine Learning Algorithm 1 Decision Tree Random Forest Algorithm 2 Algorithm 3 Algorithm 4 Linear Regression Logistic Regression
- 29. Popular Algorithms in Machine Learning Algorithm 1 Decision Tree Random Forest K Nearest Neighbors Algorithm 2 Algorithm 3 Algorithm 4 Algorithm 5 Linear Regression Logistic Regression
- 31. History of Linear Regression Linear Regression: Brief history on how Regression came into picture. Francis Galton Heights of Father and Son Studied Best fit line Regression Line Analyzed Height Data Mean Height of all People Regressed
- 32. Linear Regression Linear Regression is a linear modelling approach to find relationship between one or more independent variables (predictors) denoted as X and dependent variable (target) denoted as Y. Temperature Sales Temperature Sales Regression line to predict the sales Plotting the sales of ice cream based on temperature Predict sales of Ice Cream based on temperature:
- 33. Linear Regression Finding the best fit line: The best fit line can be found out by minimizing the distance between all the data points and the distance to the regression line. Ways to minimize this distance are sum of squared errors, sum of absolute errors etc. Distance between the data points and the line is maximum Not the best fit line
- 34. Finding the best fit line: The best fit line can be found out by minimizing the distance between all the data points and the distance to the regression line. Ways to minimize this distance are sum of squared errors, sum of absolute errors etc. Linear Regression Not the best fit line Distance between the data points and the line can still be reduced
- 35. Linear Regression Finding the best fit line: The best fit line can be found out by minimizing the distance between all the data points and the distance to the regression line. Ways to minimize this distance are sum of squared errors, sum of absolute errors etc. Distance between the data points and the line is the least Best fit regression line
- 36. Implementation of Linear Regression Linear Regression: Predict Employee Salary based on Years of Experience. HR team trying to figure out how much to pay to a new joinee? Predict Salary How much do we pay her? New Joinee Years of experience Salary Employee data of Salary based on Years of experience
- 37. Implementation of Linear Regression 1. Load the libraries: 2. Import the Dataset:
- 38. Implementation of Linear Regression 3. Visualize the data:
- 39. Implementation of Linear Regression 4. Split the data into training and testing set: 5. Fit Simple Linear Regression to the training dataset:
- 40. Implementation of Linear Regression 6. Predict the test set results:
- 41. Implementation of Linear Regression 7. Visualize the train set results:
- 42. Implementation of Linear Regression 8. Visualize the test set results:
- 43. Implementation of Linear Regression 9. Calculating the residuals: Since the RMSE value is too less, the model is a good model
- 45. Logistic Regression Logistic Regression is a Classification algorithm used to predict discrete/ categorical values Who will default on their credit card payment?
- 46. Logistic Regression Logistic Regression is a Classification algorithm used to predict discrete/ categorical values Who will default on their credit card payment? Credit card users
- 47. Logistic Regression Logistic Regression is a Classification algorithm used to predict discrete/ categorical values Who will default on their credit card payment? Credit card users Make credit card transaction
- 48. Logistic Regression Logistic Regression is a Classification algorithm used to predict discrete/ categorical values Who will default on their credit card payment? Credit card users Plot of monthly credit card balance and annual income Make credit card transaction
- 49. Logistic Regression Plotting the Logistic Regression Curve: The Logistic Regression curve is known as the Sigmoid curve (S curve) Predicted Y lies in the range 0 and 1 1 1 + e-zP = Predicted Y can exceed the range 0 and 1 Y = m1x+c0
- 50. Logistic Regression Plotting the Logistic Regression Curve: The Logistic Regression curve is known as the Sigmoid curve (S curve) Y = m1x+c0 0.5 Threshold value 0.5 Cutoff point at 0.5, anything below it results in 0 and above is 1 Red data point will default as it is above the threshold value of 0.5 and green data point won’t as it is below the threshold value
- 51. Implementation of Logistic Regression Logistic Regression: Predict if a person will buy an SUV based on their Age and Estimated Salary Age Estimated Salary
- 52. Implementation of Logistic Regression 1. Load the libraries: 2. Import the dataset and extract the independent and dependent variables:
- 53. Implementation of Logistic Regression 3. Visualize the dataset:
- 54. Implementation of Logistic Regression 4. Split the dataset into Training and Testing set: 5. Feature Scaling: 6. Fit Logistic Regression to Training dataset:
- 55. Implementation of Logistic Regression 7. Predicting the Test set results:
- 56. Implementation of Logistic Regression 8. Visualize the Train set results:
- 57. Implementation of Logistic Regression 9. Visualize the Test set results:
- 58. Implementation of Logistic Regression 11. Evaluating the model: Understanding the Confusion Matrix: N = 134 Predicted: No Predicted: Yes Actual: No TN=79 FP=6 Actual: Yes FN=11 TP=38 Accuracy: (TN+TP)/N (79+38)/134 = 0.87 Misclassification Rate: (FP+FN)/N (6+11)/134 = 0.13
- 59. Decision Tree and Random Forest
- 60. Decision Tree Decision Tree is a tree where each node represents a feature (attribute), each link (branch) represents a decision and each leaf represents an outcome (categorical or continuous value). Should the person accept a new job offer? Yes Salary > $60,000 Root Node Accept a job offer? Reject Offer No Leaf node Accept offer Yes Reject Offer No Reject Offer YesNo Performance Incentives Decision node Commute > 1 hour
- 61. Implementation of Decision Tree Decision Tree and Random Forest: Does Kyphosis exist after a surgery? Bunch of kids Kyphosis surgery Predict kyphosis present or absent?
- 62. Implementation of Decision Tree Classification Tree for Kyphosis: Start>=8.5 Absent 64/17 Present 8/11 No Start>=14.5 Absent 56/6 Yes Absent 12/0 Yes Absent 29/0 Yes Absent 12/2 Yes Age<55 Absent 27/6 No Age>=111 Absent 15/6 No Present 3/4 No
- 63. Implementation of Decision Tree 1. Load the libraries: 2. Import the dataset and extract the independent and dependent variables:
- 64. Implementation of Decision Tree 3. Visualize the data:
- 65. Implementation of Decision Tree 4. Split the data into Train and Test set: 5. Train a Decision Tree:
- 66. Implementation of Decision Tree 6. Predict the Model:
- 67. Implementation of Decision Tree N = 25 Predicted: No Predicted: Yes Actual: No TN=16 FP=4 Actual: Yes FN=2 TP=3 7. Evaluate the Model: Understanding the confusion matrix: Accuracy: (TN+TP)/N (16+3)/25 = 0.76 Misclassification Rate: (FP+FN)/N (4+2)/25 = 0.24
- 68. Improving performance using Random Forest Compare the model using Random Forest Tree: Accuracy: (TN+TP)/N (18+2)/25 = 0.80 Misclassification Rate: (FP+FN)/N (2+3)/25 = 0.20
- 69. Algorithm 5 K Nearest Neighbors
- 70. K Nearest Neighbors (KNN) K Nearest Neighbors: KNN is a Classification algorithm generally used to predict categorical values. Height(ft) Weight (lbs) Class Dog 1.5 3.5 2.0 3.0 2.5 1.0 0.5 0 65 75 9585 105 Class Cat Weight (lbs) Class Dog 1.5 3.5 2.0 3.0 2.5 1.0 0.5 0 65 75 9585 105 Height(ft) New data point Class Cat To find if a new data point is a or a ?
- 71. Choosing a K will define what class a new data point is assigned to: K Nearest Neighbors Height(ft) Class Dog Weight (lbs) 0.5 2.5 2.0 1.5 3.0 1.0 3.5 10565 75 85 95 Class Cat K=3 If K=3, the new data point belongs to class Cat K=7 If K=7, the new data point belongs to class Dog
- 72. Implementation of KNN Predict if a person will buy an SUV based on Age and Estimated Salary Age Estimated Salary
- 73. Implementation of KNN 1. Load the Libraries: 2. Load the Dataset:
- 74. Implementation of KNN 3. Visualize the dataset:
- 75. Implementation of KNN 4. Split the data into Train and Test set: 5. Feature Scaling:
- 76. Implementation of KNN 6. Fit KNN to Train set: 7. Predict the Test set results:
- 77. Implementation of KNN 8. Visualize the Train set results:
- 78. Implementation of KNN 9. Visualize the Test set results:
- 79. Implementation of KNN 10. Evaluate the model by creating a confusion matrix: Understanding the confusion matrix: N = 100 Predicted: No Predicted: Yes Actual: No TN=64 FP=4 Actual: Yes FN=3 TP=29 Accuracy: (TN+TP)/N (64+29)/100 = 0.93 Misclassification Rate: (FP+FN)/N (4+3)/100 = 0.07
- 80. Summary
Editor's Notes
- #2: Style - 01
- #3: Style - 01
- #4: Style - 01
- #5: Style - 01
- #6: Style - 01
- #7: Style - 01
- #8: Style - 01
- #9: Style - 01
- #11: Style - 01
- #13: Style - 01
- #14: Style - 01
- #15: Style - 01
- #16: Style - 01
- #17: Style - 01
- #18: Style - 01
- #19: Style - 01
- #20: Style - 01
- #21: Style - 01
- #22: Style - 01
- #23: Style - 01
- #24: Style - 01
- #25: Style - 01
- #26: Style - 01
- #27: Style - 01
- #28: Style - 01
- #29: Style - 01
- #30: Style - 01
- #31: Style - 01
- #32: Style - 01
- #33: Style - 01
- #34: Style - 01
- #35: Style - 01
- #36: Style - 01
- #37: Style - 01
- #38: Style - 01
- #39: Style - 01
- #40: Style - 01
- #41: Style - 01
- #42: Style - 01
- #43: Style - 01
- #44: Style - 01
- #45: Style - 01
- #46: Style - 01
- #47: Style - 01
- #48: Style - 01
- #49: Style - 01
- #50: Style - 01
- #51: Style - 01
- #52: Style - 01
- #53: Style - 01
- #54: Style - 01
- #55: Style - 01
- #56: Style - 01
- #57: Style - 01
- #58: Style - 01
- #59: Style - 01
- #60: Style - 01
- #61: Style - 01
- #62: Style - 01
- #63: Style - 01
- #64: Style - 01
- #65: Style - 01
- #66: Style - 01
- #67: Style - 01
- #68: Style - 01
- #69: Style - 01
- #70: Style - 01
- #71: Style - 01
- #72: Style - 01
- #73: Style - 01
- #74: Style - 01
- #75: Style - 01
- #76: Style - 01
- #77: Style - 01
- #78: Style - 01
- #79: Style - 01
- #80: Style - 01
- #81: Style - 01