Download as PDF, PPTX
Uploaded byJavier Arias Losada
Pybcn machine learning for dummies with python
The document is a presentation on machine learning (ML) with a focus on its significance, various technologies, and basic concepts, including practical examples and code samples using Python. It outlines the steps involved in the ML process, from data collection to algorithm training and prediction validation, while discussing tools like Python, sklearn, and TensorFlow. Additionally, it addresses the complexity of ML and the mathematics behind it, emphasizing the importance of accuracy and potential challenges like overfitting.
More Related Content
Similar to Pybcn machine learning for dummies with python
Pybcn machine learning for dummies with python
- 1. Machine Learning for dummies withPython PYBCN Javier Arias @javier_arilos
- 2. One day inyour life June 2016
- 3. One day inyour life Time to leave the office
- 4. One day inyour life Tesla autopilot
- 5. One day inyour life
- 6. One day inyour life Playing music
- 7. One day inyour life Your photos organized
- 8. One day inyour life Machine Learning is here, it is everywhere and it is going to stay
- 9. About this presentation WhyMachine Learning (ML) matters A journey on Machine Learning Some ML technologies and resources Some basic ML concepts, with code sample
- 10. Machine Learning isthe next big thing
- 11. Are machines alreadyintelligent?
- 12. Image-net challenge 2015: machinesoutperform people
- 13. Chess 1997: Deepblue defeatsKasparov
- 14. Game of Go 2016:AlphaGo wins world champion Lee Sedol
- 16.
- 17. Learning about ML MOOC- Massive Open Online Courses Contents by the best universities and companies Some superstar teachers Udacity, Coursera, EdX
- 18. Udacity - Introto Machine Learning Pattern Recognition for Fun and Profit - Very well organized contents - Python + sklearn - Sebastian Thrun - Free - At your own pace
- 19. Udacity - Introto Machine Learning Pattern Recognition for Fun and Profit
- 20. Udacity - Introto Machine Learning Pattern Recognition for Fun and Profit
- 21. Udacity - DeepLearning Take machine Learning to the next level ML branch based on algorithms that use multiple processing layers ● By Google ● Python and Tensorflow ● No wine for the moment :-(
- 22. What is MachineLearning?
- 23. Solving a complexproblem somethingfeatures (data) prediction
- 24. First approach, programming tellthe computer what to do in very tiny steps
- 25. … represented primarily asif-then rules ... Artificial Intelligence - expert systems
- 26. Machine Learning show thecomputer some real world data and let it learn from it
- 27. Machine Learning, implications Wecan train computers to do things we do not know how to do
- 29. ML example: characterrecognition Not-MNIST dataset Thousands of 28x28 grayscale images with labels
- 30. features x 1000slabels x 1000s F G F J ML step 1: get samples (training data)
- 31. ML step 2:choose an algorithm Linear regression Support Vector Mach. k-Means Decision Trees Random Forests Neural networks Convolutional NN Naive Bayes
- 32. ML step 3:train your algorithm features x 1000s labels x 1000sML algorithm F G F J
- 33. ML, last step:getting predictions ML algorithmfeatures (data) prediction D
- 34. Tricky Question How goodare our predictions?
- 35.
- 36. The Tools Python sklearntensorflow
- 37. The Tools: Python ●Opensource ● Expressive ● Interpreted, dynamically typed ● Widely used many different problems ● Batteries included: Notebook, Libraries
- 38. The Tools: sklearn ●Opensource, Python ● Wonderful documentation ● Support to full ML lifecycle: ○ Feauture engineering ○ Algorithms ○ Validation ○ Datasets
- 39. The Tools: TensorFlow ●Opensource, Python ● Deep Learning ● Data flow graphs. ○ Nodes: mathematical operations ○ Edges: Tensors, multidimensional arrays
- 41. A summary ofML process ● Get features (with labels) ● Choose and configure an algorithm ● Train your algorithm ● Do predictions ● Validate your results
- 42. train your model tr_ds,_, tr_lbl, _ = train_test_split(dataset, labels, train_size=size, random_state=17) clf = LogisticRegression() clf.fit(tr_ds, tr_lbl) # fit with train dataset and train labels train_ds test_ds dataset
- 43. make predictions pred =clf.predict(test_dataset) How good are our predictions?
- 44. accuracy test_predicions = clf.predict(test_dataset) acc= accuracy_score(test_labels, test_predictions)
- 45.
- 46. Improving prediction results Trainingdata Algorithm + config
- 47. Is Machine Learningblack magic?
- 48. Machine Learning ismathematics
- 49. Linear regression -predict house prices
- 50.
- 51. Linear regression price =m·size + n
- 52. Linear regression -starting training
- 53. Linear regression -training iteration
- 54. Linear regression -error minimized
- 55.
- 56. ML finds boundariesin your data
- 58. 89% accuracy… Itcan never be too much.
- 59. First trained algorithm- 30% accuracy
- 60. 75% accuracy -should we aim for 100%?
- 61. 100% accuracy, nowwe are talking!
- 62.
- 63.
- 71.