CodeLess Machine Learning
Download as PPTX, PDF0 likes347 views
This document provides an introduction to machine learning concepts including definitions of machine learning, training and test data, and different machine learning techniques. It defines machine learning as a field that allows machines to learn from data without being explicitly programmed. It describes how training data is used to teach a machine and test data is used to evaluate how well a machine has learned. The document outlines common machine learning techniques including supervised learning techniques like classification and regression as well as unsupervised learning techniques like clustering. It provides examples of different algorithms for each technique.
![References
Web sites:
[1] Machine Learning in R for Beginners, url:https://www.edureka.co/blog/machine-learning-with-r
[2] Choose your x and y carefully., https://mathbench.umd.edu/modules/visualization_graph/page02.htm
[3] Text Classification, https://monkeylearn.com/text-classification/
[4] Dataset, https://storm.cis.fordham.edu/~gweiss/data-mining/datasets.html
[5] Dataset, https://archive.ics.uci.edu/ml/datasets.php](https://image.slidesharecdn.com/160930-machine-learning-template-16x9-200309054927/85/CodeLess-Machine-Learning-44-320.jpg)
CodeLess Machine Learning
- 1. Codeless Programming Machine Learning Ibra College of Technology Information Technology Sharjeel Imtiaz Lecturer IT Under the guidance of Faiza Rashid Ammar Al-Harthy A Smart City Initiative towards Machine Learning and AI
- 2. Machine Learning • What is Machine Learning? • What is learning with train/test data?
- 3. Introduction to ML • What is ML?
- 5. Introduction to ML • Your Parents told you that this is a fish and it has some specific features associated with it like it has fins, gills, a pair of eyes, a tail and so on. Now, whenever your brain comes across an image with those set of features, it automatically registers it as a fish because your brain has learned that it is a fish. And which type.
- 6. Introduction to ML • That’s how our brain functions but what about a machine? If the same image is fed to a machine, how will the machine identify it to be a fish?
- 7. Introduction to ML • his is where Machine Learning comes in. We’ll keep on feeding images of a fish to a computer with the tag “fish” until the machine learns all the features associated with a fish.
- 8. Introduction to ML • What is training data / test data? • Once the machine learns all the features associated with a fish, we will feed it new data to determine how much has it learned.
- 9. Introduction to ML • What is training data / test data? • In other words, Raw Data/Training Data is given to the machine, so that it learns all the features associated with the Training Data. Once, the learning is done, it is given New Data/Test Data to determine how well the machine has learned.
- 10. Types of Machine Learning Techniques • Supervised Learning 1. Classification 2. Regression • Unsupervised Learning
- 11. Types of Machine Learning Techniques • Classification (Train/Test Data) Determines to which set of categories does a new observation belongs. 1. A classification algorithm learns all the features and labels of the training data and 2. when new data is given to it, it has to assign labels to the new observations depending on what it has learned from the training data.
- 12. Types of Machine Learning Techniques • Classification (wright / wrong classes)
- 13. Types of Machine Learning Techniques Classification Techniques sentiment analysis, topic labeling, language detection, and intent detection. Sentiment Analysis Probably the most common example of text classification is sentiment analysis: the automated process of determining whether a text is positive, negative, or neutral
- 14. Types of Machine Learning Techniques • SVM • Naïve Bayesian • Neural network • Decision Tree or recursive tree
- 15. Types of Machine Learning Techniques • Example: • Conducted survey to see what customers were interested in new model car • Want to select customers for advertising campaign sale custId car age city newCar c1 taurus 27 sf yes c2 van 35 la yes c3 van 40 sf yes c4 taurus 22 sf yes c5 merc 50 la no c6 taurus 25 la no
- 17. Neural network • Take input in neurons or tensors • Dense connectivity at hidden • Output layer used softmax function • Hidden layers using linear functions or sigmoid
- 18. SVM • Best line separate the tow classes • See the maximum gap between hyperplane
- 19. K-Means Clustering K-mean • Take mean and close points group together • Grouping can done in clusters or hierarchies
- 20. Types of Machine Learning Techniques • Classification (Train/Test Data) Short Texts: Text classification can be used in a broad range of contexts such as classifying short texts (e.g. as tweets, headlines or tweets). Larger Documents: Organizing much larger documents (e.g. customer reviews, media articles or legal contracts).
- 21. Types of Machine Learning Techniques • Open text and input in above model amazon_review.txt • https://app.monkeylearn.com/main/classifiers/cl_pi3C7JiL/
- 22. Types of Machine Learning Techniques Topic Modeling • Another common example of text classification is topic labeling, that is, understanding what a given text is talking about. • It’s often used for structuring and organizing data such as organizing customer feedback by its topic or organizing news articles according to their subject.
- 23. Types of Machine Learning Techniques Topic Modeling • Open text and input in above model amazon_review.txt • https://app.monkeylearn.com/main/classifiers/cl_sGdE8hD9/
- 24. Types of Machine Learning Techniques Language Detection Language detection is another great example of text classification, that is, the process of classifying incoming text according to its language • Open text and input in above model amazon_review.txt • https://app.monkeylearn.com/main/classifiers/cl_Vay9jh28/
- 25. Types of Machine Learning Techniques Intent Detection Companies are also using text classifiers for automatically detecting the intent from customer conversations, often used for generating product analytics or automating business purposes. • Open text and input in above model amazon_review.txt • https://app.monkeylearn.com/main/classifiers/cl_v9GTn7zi/
- 26. Types of Machine Learning Techniques • Regression / Prediction • Regression is a supervised learning algorithm which helps in determining how does one variable influence another variable. • Regression (LM) • Neural Network
- 27. Types of Machine Learning Techniques • Supervised Learning 1. Classification 2. Regression • Unsupervised Learning
- 28. Types of Machine Learning Techniques • Regression / Prediction (Dependent variable/ Independent) • living_area” is the independent variable and “price” is the dependent variable how does “price” vary with respect to “living_area
- 29. Types of Machine Learning Techniques • Refer to Data Visualization Sheet 1 • https://mathbench.umd.edu/modules/visualization_ graph/page02.htm
- 30. Types of Machine Learning Techniques • Case Study: Brand Monitoring The online conversation around a brand and its competitors heavily influences consumers. Some blogs, forums, review sites, and influencers are becoming more important than traditional outlets.
- 31. Types of Machine Learning Techniques • Unsupervised Learning: Unsupervised learning algorithm draws inferences from data which does not have labels. Clustering is an example of unsupervised learning. • “K-means” • “Hierarchical” • “Fuzzy C-Means” are some examples of clustering algorithms.
- 32. Types of Machine Learning Techniques • Similarity form the cluster • Intra cluster similarity • Inter cluster similarity • High intra cluster Similarity • Low inter cluster Similarity (why)
- 33. Interactive K-Mean Clustering Calculator https://people.revoledu.com/kardi/tutorial/kMean/Online-K-Means-Clustering.html Data st,age,cgpa,level A,20,3.3,1 B,10,2.0,2 C,11,2.1,1 D,11,2.2,1 sheetslevel.txt Types of Machine Learning Techniques
- 34. The following methods for validation will be demonstrated: • Train/test split criteria 80/20 • k-Fold Cross-Validation 5 equal dataset 20% each • Leave-one-out Cross-Validation • Leave-one-group-out Cross-Validation • Nested Cross-Validation • Time-series Cross-Validation • Wilcoxon signed-rank test • McNemar’s test • 5x2CV paired t-test • 5x2CV combined F test Validation Techniques
- 35. The following methods for validation will be demonstrated: • Train/test split criteria 80/20 Over fit • k-Fold Cross-Validation 5 equal dataset 20% each • does this for all combinations and averages the result on each instance. Validation Techniques
- 36. The following methods for validation will be demonstrated: • Train/test split criteria 80/20 • Leave-one-out Cross-Validation (LOOCV) 10 equal sample 10% each but not combined accuracy each set is separate test Validation Techniques
- 37. • Small example Diabetic Patient Case Study Number of Instances: 768 Number of Attributes: 8 plus class For Each Attribute: (all numeric-valued) 1. Number of times pregnant 2. Plasma glucose concentration a 2 hours in an oral glucose tolerance test 3. Diastolic blood pressure (mm Hg) 4. Triceps skin fold thickness (mm) 5. 2-Hour serum insulin (mu U/ml) 6. Body mass index (weight in kg/(height in m)^2) 7. Diabetes pedigree function 8. Age (years) 9. Class variable (0 or 1)
- 38. • Small example Diabetic Patient • Teacher will give demo in Weka • Student in group see the results Case Study1
- 39. • Weather forecasting • Teacher will give demo in Weka • Student in group see the results Case Study2 Demo
- 40. • Iris • Teacher will give demo in Weka • Student in group see the results Case Study3 Demo
- 41. • Environment • Health • Tourism • Geographical (map) • Biological domains area
- 42. Relational Data (Tables/Transaction/Legacy Data) - SQL Lite Text Data (Web) - Unstructured Text, csv, JSON Semi-structured Data (XML) Graph Data Social Network, Semantic Web (RDF), … Streaming Data You can only scan the data once Type of data
- 43. NLP my Interest and latest trend • SENTIMENT analysis : analyze sentiment and emtions To cater mass data (100million reviews) using map-reduce • Topic analysis : analyze amazon review topic for new rating criteria • Social networking : facebook or linkedin or analyze and make graph which produce million nodes for genome
- 44. References Web sites: [1] Machine Learning in R for Beginners, url:https://www.edureka.co/blog/machine-learning-with-r [2] Choose your x and y carefully., https://mathbench.umd.edu/modules/visualization_graph/page02.htm [3] Text Classification, https://monkeylearn.com/text-classification/ [4] Dataset, https://storm.cis.fordham.edu/~gweiss/data-mining/datasets.html [5] Dataset, https://archive.ics.uci.edu/ml/datasets.php