Python Machine Learning Cookbook: Over 100 recipes to progress from smart data analytics to deep learning using real-world datasets (2nd Edition)

Description

This eagerly anticipated second edition of the popular Python Machine Learning Cookbook will enable you to adopt a fresh approach to dealing with real-world machine learning and deep learning tasks.

With the help of over 100 recipes, you will learn to build powerful machine learning applications using modern libraries from the Python ecosystem. The book will also guide you on how to implement various machine learning algorithms for classification, clustering, and recommendation engines, using a recipe-based approach. With emphasis on practical solutions, dedicated sections in the book will help you to apply supervised and unsupervised learning techniques to real-world problems. Toward the concluding chapters, you will get to grips with recipes that teach you advanced techniques including reinforcement learning, deep neural networks, and automated machine learning.

By the end of this book, you will be equipped with the skills you need to apply machine learning techniques and leverage the full capabilities of the Python ecosystem through real-world examples.

What you will learn

• Use predictive modeling and apply it to real-world problems
• Explore data visualization techniques to interact with your data
• Learn how to build a recommendation engine
• Understand how to interact with text data and build models to analyze it
• Work with speech data and recognize spoken words using Hidden Markov Models
• Get well versed with reinforcement learning, automated ML, and transfer learning
• Work with image data and build systems for image recognition and biometric face recognition
• Use deep neural networks to build an optical character recognition system

Who this book is for

This book is for data scientists, machine learning developers, deep learning enthusiasts and Python programmers who want to solve real-world challenges using machine-learning techniques and algorithms. If you are facing challenges at work and want ready-to-use code solutions to cover key tasks in machine learning and the deep learning domain, then this book is what you need. Familiarity with Python programming and machine learning concepts will be useful.

Key Features

• Learn and implement machine learning algorithms in a variety of real-life scenarios
• Cover a range of tasks catering to supervised, unsupervised and reinforcement learning techniques
• Find easy-to-follow code solutions for tackling common and not-so-common challenges

Table of Contents

1. Title Page
2. Copyright and Credits
3. Python Machine Learning Cookbook Second Edition
4. About Packt
5. Why subscribe?
6. Contributors
7. About the author
8. About the reviewer
9. Packt is searching for authors like you
10. Preface
11. Who this book is for
12. What this book covers
13. To get the most out of this book
14. Download the example code files
15. Download the color images
16. Conventions used
17. Sections
18. Getting ready
19. How to do it…
20. How it works…
21. There’s more…
22. See also
23. Get in touch
24. Reviews
25. The Realm of Supervised Learning
26. Technical requirements
27. Introduction
28. Array creation in Python
29. Getting ready
30. How to do it…
31. How it works…
32. There’s more…
33. See also
34. Data preprocessing using mean removal
35. Getting ready
36. How to do it…
37. How it works…
38. There’s more…
39. See also
40. Data scaling
41. Getting ready
42. How to do it…
43. How it works…
45. There’s more…
46. See also
47. Normalization
48. Getting ready
49. How to do it…
50. How it works…
51. There’s more…
52. See also
53. Binarization
54. Getting ready
55. How to do it…
56. How it works…
57. There’s more…
58. See also
59. One-hot encoding
60. Getting ready
61. How to do it…
62. How it works…
63. There’s more…
64. See also
65. Label encoding
66. Getting ready
67. How to do it…
68. How it works…
69. There’s more…
70. See also
71. Building a linear regressor
72. Getting ready
73. How to do it…
74. How it works…
75. There’s more…
76. See also
77. Computing regression accuracy
78. Getting ready
79. How to do it…
80. How it works…
81. There’s more…
82. See also
83. Achieving model persistence
84. Getting ready
85. How to do it…
86. How it works…
87. There’s more…
88. See also
89. Building a ridge regressor
90. Getting ready
91. How to do it…
92. How it works…
93. See also
94. Building a polynomial regressor
95. Getting ready
96. How to do it…
97. How it works…
98. There’s more…
99. See also
100. Estimating housing prices
101. Getting ready
102. How to do it…
103. How it works…
104. There’s more…
105. See also
106. Computing the relative importance of features
107. Getting ready
108. How to do it…
109. How it works…
110. There’s more…
111. See also
112. Estimating bicycle demand distribution
113. Getting ready
114. How to do it…
115. How it works…
116. There’s more…
117. See also
118. Constructing a Classifier
119. Technical requirements
120. Introduction
121. Building a simple classifier
122. Getting ready
123. How to do it…
124. How it works…
125. There’s more…
126. See also
127. Building a logistic regression classifier
128. Getting ready
129. How to do it…
130. How it works…
131. There’s more…
132. See also
133. Building a Naive Bayes classifier
134. Getting ready
135. How to do it…
136. How it works…
137. There’s more…
138. See also
139. Splitting a dataset for training and testing
140. Getting ready
141. How to do it…
142. How it works…
143. There’s more…
144. See also
145. Evaluating accuracy using cross-validation metrics
146. Getting ready…
147. How to do it…
148. How it works…
149. There’s more…
150. See also
151. Visualizing a confusion matrix
152. Getting ready
153. How to do it…
154. How it works…
155. There’s more…
156. See also
157. Extracting a performance report
158. Getting ready
159. How to do it…
160. How it works…
161. There’s more…
162. See also
163. Evaluating cars based on their characteristics
164. Getting ready
165. How to do it…
166. How it works…
167. There’s more…
168. See also
169. Extracting validation curves
170. Getting ready
171. How to do it…
172. How it works…
173. There’s more…
174. See also
175. Extracting learning curves
176. Getting ready
177. How to do it…
178. How it works…
179. There’s more…
180. See also
181. Estimating the income bracket
182. Getting ready
183. How to do it…
184. How it works…
185. There’s more…
186. See also
187. Predicting the quality of wine
188. Getting ready
189. How to do it…
190. How it works…
191. There’s more…
192. See also
193. Newsgroup trending topics classification
194. Getting ready
195. How to do it…
196. How it works…
197. There’s more…
198. See also
199. Predictive Modeling
200. Technical requirements
201. Introduction
202. Building a linear classifier using SVMs
203. Getting ready
204. How to do it…
205. How it works…
206. There’s more…
207. See also
208. Building a nonlinear classifier using SVMs
209. Getting ready
210. How to do it…
211. How it works…
212. There’s more…
213. See also
214. Tackling class imbalance
215. Getting ready
216. How to do it…
217. How it works…
218. There’s more…
219. See also
220. Extracting confidence measurements
221. Getting ready
222. How to do it…
223. How it works…
224. There’s more…
225. See also
226. Finding optimal hyperparameters
227. Getting ready
228. How to do it…
229. How it works…
230. The grid search algorithm
231. The randomized search algorithm
232. The Bayesian optimization algorithm
233. There’s more…
234. See also
235. Building an event predictor
236. Getting ready
237. How to do it…
238. How it works…
239. There’s more…
240. See also
241. Estimating traffic
242. Getting ready
243. How to do it…
244. How it works…
245. There’s more…
246. See also
247. Simplifying machine learning workflow using TensorFlow
248. Getting ready
249. How to do it…
250. How it works…
251. There’s more…
252. See also
253. Implementing a stacking method
254. Getting ready
255. How to do it…
256. How it works…
257. There’s more…
258. See also
259. Clustering with Unsupervised Learning
260. Technical requirements
261. Introduction
262. Clustering data using the k-means algorithm
263. Getting ready
264. How to do it…
265. How it works…
266. There’s more…
267. See also
268. Compressing an image using vector quantization
269. Getting ready
270. How to do it…
271. How it works…
272. There’s more…
273. See also
274. Grouping data using agglomerative clustering
275. Getting ready
276. How to do it…
277. How it works…
278. There’s more…
279. See also
280. Evaluating the performance of clustering algorithms
281. Getting ready
282. How to do it…
283. How it works…
284. There’s more…
285. See also
286. Estimating the number of clusters using the DBSCAN algorithm
287. Getting ready
288. How to do it…
289. How it works…
290. There’s more…
291. See also
292. Finding patterns in stock market data
293. Getting ready
294. How to do it…
295. How it works…
296. There’s more…
297. See also
298. Building a customer segmentation model
299. Getting ready
300. How to do it…
301. How it works…
302. There’s more…
303. See also
304. Using autoencoders to reconstruct handwritten digit images
305. Getting ready
306. How to do it…
307. How it works…
308. There’s more…
309. See also
310. Visualizing Data
311. Technical requirements
312. An introduction to data visualization
313. Plotting three-dimensional scatter plots
314. Getting ready
315. How to do it…
316. How it works…
317. There’s more…
318. See also
319. Plotting bubble plots
320. Getting ready
321. How to do it…
322. How it works…
323. There’s more…
324. See also
325. Animating bubble plots
326. Getting ready
327. How to do it…
328. How it works…
329. There’s more…
330. See also
331. Drawing pie charts
332. Getting ready
333. How to do it…
334. How it works…
335. There’s more…
336. See also
337. Plotting date-formatted time series data
338. Getting ready
339. How to do it…
340. How it works…
341. There’s more…
342. See also
343. Plotting histograms
344. Getting ready
345. How to do it…
346. How it works…
347. There’s more…
348. See also
349. Visualizing heat maps
350. Getting ready
351. How to do it…
352. How it works…
353. There’s more…
354. See also
355. Animating dynamic signals
356. Getting ready
357. How to do it…
358. How it works…
359. There’s more…
360. See also
361. Working with the Seaborn library
362. Getting ready
363. How to do it…
364. How it works…
365. There’s more…
366. See also
367. Building Recommendation Engines
368. Technical requirements
369. Introducing the recommendation engine
370. Building function compositions for data processing
371. Getting ready
372. How to do it…
373. How it works…
374. There’s more…
375. See also
376. Building machine learning pipelines
377. Getting ready
378. How to do it…
379. How it works…
380. There’s more…
381. See also
382. Finding the nearest neighbors
383. Getting ready
384. How to do it…
385. How it works…
386. There’s more…
387. See also
388. Constructing a k-nearest neighbors classifier
389. Getting ready
390. How to do it…
391. How it works…
392. There’s more…
393. See also
394. Constructing a k-nearest neighbors regressor
395. Getting ready
396. How to do it…
397. How it works…
398. There’s more…
399. See also
400. Computing the Euclidean distance score
401. Getting ready
402. How to do it…
403. How it works…
404. There’s more…
405. See also
406. Computing the Pearson correlation score
407. Getting ready
408. How to do it…
409. How it works…
410. There’s more…
411. See also
412. Finding similar users in the dataset
413. Getting ready
414. How to do it…
415. How it works…
416. There’s more…
417. See also
418. Generating movie recommendations
419. Getting ready
420. How to do it…
421. How it works…
422. There’s more…
423. See also
424. Implementing ranking algorithms
425. Getting ready
426. How to do it…
427. How it works…
428. There’s more…
429. See also
430. Building a filtering model using TensorFlow
431. Getting ready
432. How to do it…
433. How it works…
434. There’s more…
435. See also
436. Analyzing Text Data
437. Technical requirements
438. Introduction
439. Preprocessing data using tokenization
440. Getting ready
441. How to do it…
442. How it works…
443. There’s more…
444. See also
445. Stemming text data
446. Getting ready
447. How to do it…
448. How it works…
449. There’s more…
450. See also
451. Converting text to its base form using lemmatization
452. Getting ready
453. How to do it…
454. How it works…
455. There’s more…
456. See also
457. Dividing text using chunking
458. How to do it…
459. How it works…
460. There’s more…
461. See also
462. Building a bag-of-words model
463. Getting ready
464. How to do it…
465. How it works…
466. There’s more…
467. See also
468. Building a text classifier
469. Getting ready
470. How to do it…
471. How it works…
472. There’s more…
473. See also
474. Identifying the gender of a name
475. Getting ready
476. How to do it…
477. How it works…
478. There’s more…
479. See also
480. Analyzing the sentiment of a sentence
481. Getting ready
482. How to do it…
483. How it works…
484. There’s more…
485. See also
486. Identifying patterns in text using topic modeling
487. Getting ready
488. How to do it…
489. How it works…
490. There’s more…
491. See also
492. Parts of speech tagging with spaCy
493. Getting ready
494. How to do it…
495. How it works…
496. There’s more…
497. See also
498. Word2Vec using gensim
499. Getting ready
500. How to do it…
501. How it works…
502. There’s more…
503. See also
504. Shallow learning for spam detection
505. Getting ready
506. How to do it…
507. How it works…
508. There’s more…
509. See also
510. Speech Recognition
511. Technical requirements
512. Introducing speech recognition
513. Reading and plotting audio data
514. Getting ready
515. How to do it…
516. How it works…
517. There’s more…
518. See also
519. Transforming audio signals into the frequency domain
520. Getting ready
521. How to do it…
522. How it works…
523. There’s more…
524. See also
525. Generating audio signals with custom parameters
526. Getting ready
527. How to do it…
528. How it works…
529. There’s more…
530. See also
531. Synthesizing music
532. Getting ready
533. How to do it…
534. How it works…
535. There’s more…
536. See also
537. Extracting frequency domain features
538. Getting ready
539. How to do it…
540. How it works…
541. There’s more…
542. See also
543. Building HMMs
544. Getting ready
545. How to do it…
546. How it works…
547. There’s more…
548. See also
549. Building a speech recognizer
550. Getting ready
551. How to do it…
552. How it works…
553. There’s more…
554. See also
555. Building a TTS system
556. Getting ready
557. How to do it…
558. How it works…
559. There’s more…
560. See also
561. Dissecting Time Series and Sequential Data
562. Technical requirements
563. Introducing time series
564. Transforming data into a time series format
565. Getting ready
566. How to do it…
567. How it works…
568. There’s more…
569. See also
570. Slicing time series data
571. Getting ready
572. How to do it…
573. How it works…
574. There’s more…
575. See also
576. Operating on time series data
577. Getting ready
578. How to do it…
579. How it works…
580. There’s more…
581. See also
582. Extracting statistics from time series data
583. Getting ready
584. How to do it…
585. How it works…
586. There’s more…
587. See also
588. Building HMMs for sequential data
589. Getting ready
590. How to do it…
591. How it works…
592. There’s more…
593. See also
594. Building CRFs for sequential text data
595. Getting ready
596. How to do it…
597. How it works…
598. There’s more…
599. See also
600. Analyzing stock market data
601. Getting ready
602. How to do it…
603. How it works…
604. There’s more…
605. See also
606. Using RNNs to predict time series data
607. Getting ready
608. How to do it…
609. How it works…
610. There’s more…
611. See also
612. Analyzing Image Content
613. Technical requirements
614. Introducing computer vision
615. Operating on images using OpenCV-Python
616. Getting ready
617. How to do it…
618. How it works…
619. There’s more…
620. See also
621. Detecting edges
622. Getting ready
623. How to do it…
624. How it works…
625. There’s more…
626. See also
627. Histogram equalization
628. Getting ready
629. How to do it…
630. How it works…
631. There’s more…
632. See also
633. Detecting corners
634. Getting ready
635. How to do it…
636. How it works…
637. There’s more…
638. See also
639. Detecting SIFT feature points
640. Getting ready
641. How to do it…
642. How it works…
643. There’s more…
644. See also
645. Building a Star feature detector
646. Getting ready
647. How to do it…
648. How it works…
649. There’s more…
650. See also
651. Creating features using Visual Codebook and vector quantization
652. Getting ready
653. How to do it…
654. How it works…
655. There’s more…
656. See also
657. Training an image classifier using Extremely Random Forests
658. Getting ready
659. How to do it…
660. How it works…
661. There’s more…
662. See also
663. Building an object recognizer
664. Getting ready
665. How to do it…
666. How it works…
667. There’s more…
668. See also
669. Using Light GBM for image classification
670. Getting ready
671. How to do it…
672. How it works…
673. There’s more…
674. See also
675. Biometric Face Recognition
676. Technical requirements
677. Introduction
678. Capturing and processing video from a webcam
679. Getting ready
680. How to do it…
681. How it works…
682. There’s more…
683. See also
684. Building a face detector using Haar cascades
685. Getting ready
686. How to do it…
687. How it works…
688. There’s more…
689. See also
690. Building eye and nose detectors
691. Getting ready
692. How to do it…
693. How it works…
694. There’s more…
695. See also
696. Performing principal component analysis
697. Getting ready
698. How to do it…
699. How it works…
700. There’s more…
701. See also
702. Performing kernel principal component analysis
703. Getting ready
704. How to do it…
705. How it works…
706. There’s more…
707. See also
708. Performing blind source separation
709. Getting ready
710. How to do it…
711. How it works…
712. There’s more…
713. See also
714. Building a face recognizer using a local binary patterns histogram
715. Getting ready
716. How to do it…
717. How it works…
718. There’s more…
719. See also
720. Recognizing faces using the HOG-based model
721. Getting ready
722. How to do it…
723. How it works…
724. There’s more…
725. See also
726. Facial landmark recognition
727. Getting ready
728. How to do it…
729. How it works…
730. There’s more…
731. See also
732. User authentication by face recognition
733. Getting ready
734. How to do it…
735. How it works…
736. There’s more…
737. See also
738. Reinforcement Learning Techniques
739. Technical requirements
740. Introduction
741. Weather forecasting with MDP
742. Getting ready
743. How to do it…
744. How it works…
745. There’s more…
746. See also
747. Optimizing a financial portfolio using DP
748. Getting ready
749. How to do it…
750. How it works…
751. There’s more…
752. See also
753. Finding the shortest path
754. Getting ready
755. How to do it…
756. How it works…
757. There’s more…
758. See also
759. Deciding the discount factor using Q-learning
760. Getting ready
761. How to do it…
762. How it works…
763. There’s more…
764. See also
765. Implementing the deep Q-learning algorithm
766. Getting ready
767. How to do it…
768. How it works…
769. There’s more…
770. See also
771. Developing an AI-based dynamic modeling system
772. Getting ready
773. How to do it…
774. How it works…
775. There’s more…
776. See also
777. Deep reinforcement learning with double Q-learning
778. Getting ready
779. How to do it…
780. How it works…
781. There’s more…
782. See also
783. Deep Q-network algorithm with dueling Q-learning
784. Getting ready
785. How to do it…
786. How it works…
787. There’s more…
788. See also
789. Deep Neural Networks
790. Technical requirements
791. Introduction
792. Building a perceptron
793. Getting ready
794. How to do it…
795. How it works…
796. There’s more…
797. See also
798. Building a single layer neural network
799. Getting ready
800. How to do it…
801. How it works…
802. There’s more…
803. See also
804. Building a deep neural network
805. Getting ready
806. How to do it…
807. How it works…
808. There’s more…
809. See also
810. Creating a vector quantizer
811. Getting ready
812. How to do it…
813. How it works…
814. There’s more…
815. See also
816. Building a recurrent neural network for sequential data analysis
817. Getting ready
818. How to do it…
819. How it works…
820. There’s more…
821. See also
822. Visualizing the characters in an OCR database
823. Getting ready
824. How to do it…
825. How it works…
826. There’s more…
827. See also
828. Building an optical character recognizer using neural networks
829. Getting ready
830. How to do it…
831. How it works…
832. There’s more…
833. See also
834. Implementing optimization algorithms in ANN
835. Getting ready
836. How to do it…
837. How it works…
838. There’s more…
839. See also
840. Unsupervised Representation Learning
841. Technical requirements
842. Introduction
843. Using denoising autoencoders to detect fraudulent transactions
844. Getting ready
845. How to do it…
846. How it works…
847. There’s more…
848. See also
849. Generating word embeddings using CBOW and skipgram representations
850. Getting ready
851. How to do it…
852. How it works…
853. There’s more…
854. See also
855. Visualizing the MNIST dataset using PCA and t-SNE
856. Getting ready
857. How to do it…
858. How it works…
859. There’s more…
860. See also
861. Using word embedding for Twitter sentiment analysis
862. Getting ready
863. How to do it…
864. How it works…
865. There’s more…
866. See also
867. Implementing LDA with scikit-learn
868. Getting ready
869. How to do it…
870. How it works…
871. There’s more…
872. See also
873. Using LDA to classify text documents
874. Getting ready
875. How to do it…
876. How it works…
877. There’s more…
878. See also
879. Preparing data for LDA
880. Getting ready
881. How to do it…
882. How it works…
883. There’s more…
884. See also
885. Automated Machine Learning and Transfer Learning
886. Technical requirements
887. Introduction
888. Working with Auto-WEKA
889. Getting ready
890. How to do it…
891. How it works…
892. There’s more…
893. See also
894. Using AutoML to generate machine learning pipelines with TPOT
895. Getting ready
896. How to do it…
897. How it works…
898. There’s more…
899. See also
900. Working with Auto-Keras
901. Getting ready
902. How to do it…
903. How it works…
904. There’s more…
905. See also
906. Working with auto-sklearn
907. Getting ready
908. How to do it…
909. How it works…
910. There’s more…
911. See also
912. Using MLBox for selection and leak detection
913. Getting ready
914. How to do it…
915. How it works…
916. There’s more…
917. See also
918. Convolutional neural networks with transfer learning
919. Getting ready
920. How to do it…
921. How it works…
922. There’s more…
923. See also
924. Transfer learning with pretrained image classifiers using ResNet-50
925. Getting ready
926. How to do it…
927. How it works…
928. There’s more…
929. See also
930. Transfer learning using feature extraction with the VGG16 model
931. Getting ready
932. How to do it…
933. How it works…
934. There’s more…
935. See also
936. Transfer learning with pretrained GloVe embedding
937. Getting ready
938. How to do it…
939. How it works…
940. There’s more…
941. Inductive transfer learning
942. Unsupervised transfer learning
943. Transductive transfer learning
944. Instance transfer learning
945. See also
946. Unlocking Production Issues
947. Technical requirements
948. Introduction
949. Handling unstructured data
950. Getting ready
951. How to do it…
952. How it works…
953. There’s more…
954. See also
955. Deploying machine learning models
956. Getting ready
957. How to do it…
958. How it works…
959. There’s more…
960. See also
961. Keeping track of changes into production
962. How to do it…
963. How it works…
964. There’s more…
965. See also
966. Tracking accuracy to optimize model scaling
967. How to do it…
968. How it works…
969. There’s more…
970. See also
971. Other Books You May Enjoy
972. Leave a review – let other readers know what you think

Authors Biography

Giuseppe Ciaburro holds a PhD in environmental technical physics, along with two master’s degrees. His research was focused on machine learning applications in the study of urban sound environments. He works at the Built Environment Control Laboratory at the Universita degli Studi della Campania Luigi Vanvitelli, Italy. He has over 15 years’ professional experience in programming (Python, R, and MATLAB), first in the field of combustion, and then in acoustics and noise control. He has several publications to his credit.

Prateek Joshi is an artificial intelligence researcher, an author of several books, and a TEDx speaker. He has been featured in Forbes 30 Under 30, CNBC, TechCrunch, Silicon Valley Business Journal, and many more publications. He is the founder of Pluto AI, a venture-funded Silicon Valley start-up building an intelligence platform for water facilities. He graduated from the University of Southern California with a Master’s degree specializing in Artificial Intelligence. He has previously worked at NVIDIA and Microsoft Research.