I hope you enjoyed this last chapter of an exciting tour of machine learning. Throughout this book, we covered all of the essential topics that this field has to offer, and you should now be well equipped to put those techniques into action to solve real-world problems.

We started our journey with a brief overview of the different types of learning tasks: supervised learning, reinforcement learning, and unsupervised learning. We discussed several different learning algorithms that can be used for classification, starting with simple single-layer neural networks in Chapter 2, Training Machine Learning Algorithms for Classification. Then, we discussed more advanced classification algorithms in Chapter 3, A Tour of Machine Learning Classifiers Using Scikit-learn, and you learned about the most important aspects of a machine learning pipeline in Chapter 4, Building Good Training Sets – Data Preprocessing and Chapter 5, Compressing Data via Dimensionality Reduction. Remember that even the most advanced algorithm is limited by the information in the training data that it gets to learn from. In Chapter 6, Learning Best Practices for Model Evaluation and Hyperparameter Tuning, you learned about the best practices to build and evaluate predictive models, which is another important aspect in machine learning applications. If one single learning algorithm does not achieve the performance we desire, it can sometimes be helpful to create an ensemble of experts to make a prediction. We discussed this in Chapter 7, Combining Different Models for Ensemble Learning. In Chapter 8, Applying Machine Learning to Sentiment Analysis, we applied machine learning to analyze the probably most interesting form of data in the modern age that is dominated by social media platforms on the Internet: text documents. However, machine learning techniques are not limited to offline data analysis, and in Chapter 9, Embedding a Machine Learning Model into a Web Application, we saw how to embed a machine learning model into a web application to share it with the outside world. For the most part, our focus was on algorithms for classification, probably the most popular application of machine learning. However, this is not where it ends! In Chapter 10, Predicting Continuous Target Variables with Regression Analysis, we explored several algorithms for regression analysis to predict continuous-valued output values. Another exciting subfield of machine learning is clustering analysis, which can help us to find hidden structures in data even if our training data does not come with the right answers to learn from. We discussed this in Chapter 11, Working with Unlabeled Data – Clustering Analysis.

In the last two chapters of this book, we caught a glimpse of the most beautiful and most exciting algorithms in the whole machine learning field: artificial neural networks. Although deep learning really is beyond the scope of this book, I hope I could at least kindle your interest to follow the most recent advancement in this field. If you are considering a career as a machine learning researcher, or even if you just want to keep up to date with the current advancement in this field, I can recommend you to follow the works of the leading experts in this field, such as Geoff Hinton (http://www.cs.toronto.edu/~hinton/), Andrew Ng (http://www.andrewng.org), Yann LeCun (http://yann.lecun.com), Juergen Schmidhuber (http://people.idsia.ch/~juergen/), and Yoshua Bengio (http://www.iro.umontreal.ca/~bengioy), just to name a few. Also, please do not hesitate to join the scikit-learn, Theano, and Keras mailing lists to participate in interesting discussions around these libraries, and machine learning in general. I am looking forward to meeting you there! You are always welcome to contact me if you have any questions about this book or need some general tips about machine learning.

I hope this journey through the different aspects of machine learning was really worthwhile, and you learned many new and useful skills to advance your career and apply them to real-world problem solving.