FireShot Capture 003 - Top 12 Data Science Books That Will Boost Your Career in 2020

You might also like

You are on page 1of 1
DataOptimal Courses About Blog Contact Top 12 Data Science Books That Will ere) Boost Your Career In 2020 PP eg Learning data science can be overwhelming Theres hundreds of books, online courses, an grate degrees, Where do you start? Instead of eying o gure it ut on your ot, use tis Ist of fre data selonce textbooks. ‘Bonus: Download a feo summary shoot wth ll 12 textbooks fo start learning Immediately feo Thistncluces everything from the bascs of Python and, to advanced techniques in ‘machine learning, data mining, and statistics, Now fm not tling you to go off and read each one cover to cover. Thats not an effective strategy Instead, use these toxthooks to learn a new skill quickly Tha bert way te do that iy cins 9 pve Bling projects san effective srateny forthe folowing two reasons +. terinfoces what your learning by 2pjing the sls mediately 2 Fronds you wth portfolio projet to shoncase potential empoyers Use the iste provided below tolearn some new data scence hls and bl porto projets 1. Python Data Science Handbook Bp ake Vandertas on Data Science. Handbook Recent data shows that Python iss heeading a8 fo a selonce and machine learning ‘The Python Data scence Handbook she perfect reference for boosting your Python sls ‘Asa ata scientist yout often be asked 1 work on numerous tasts, but a majorly of your tine will be spent on manipulating data and 312 :ianins Thistsa perfect reference to heep close by for those Frequent data manipulation asks using Pandas Here's number oF other important data science topics this book covers + wytnon set 1 Numpy for computations + Data manipulation wth Pandas + Data visualizations wih Maipli> + Machin learning with Sake Learn ‘ation step: Use the data manipulaon section with Pandas ocean a messy dataset. Here’ agren place for yout find messy 01> to work ath, 2. Think Python By: Allen 8 Downey youre jst starting out programming wth Python this book for you. you a mere advanced Python user. this book sao fer you ‘Think Python reviews everything from the basis of data structares and functions to more ance topes uch a clasas ane inher, Every fe chapters this book es together kay concepts with ase studs. This great way to reinforce learning nen concepts Heres of ust feu af the topics covered inthis bake + Functions eration + Data structures + Flos + cases + Methods + tnhertance Action step: Work through th case study n Chapter 13.0n data suucture selection. Fip back and fet tothe previous chapters 3s needed, but don't read them end tend This case study a great example of how to complete a word frequency analysis, 3. R for Data Science By: Garret Grolemund and Haley Wickham R S Ifyou want to make yourself marketable to employers and stay current ith your data scence sll, you should have a good handle on Rismeckin neck with Python asthe top programming languages for data scence ‘voce pl ofthe data science community indicated that 52.1% of responders use Rony Sgt lss than 52.6% which use Python, you want to sharpen your R sil, for Data Science isthe perfect book, Iccovers the bases for new R users, sucha ot ler but aso gets into more advanced topics as wel ‘ata ster can spend up ta 805 of el ne Jong sb Hs sa relerence you wil dete want to keep cose by. Thisbookiea great general reference from Hadley Wickham and Garret Goleman twa of the top developers in the 8 community. Here's a numberof topics covered + Exploration + wrangling + Programming + Mosoing + communication Aetion Step: Use): pr to perform an expocarory analysis You can explore this sda" and decumar your fingsusing.anRrarksown notebook. "Make sure you put your project on your pv 80 and ink oe from the projects section on yourlinkedin profile 4. Advanced R By: Hadley Wickham Troy you realy want to set yours apart as an user and impress employes, Advancad Risa rat resource Iteovers everything from the foundations, Including data structures object oriented eogramming. and debugging to functional programming nd perermance code With the development ofthe Repp package, Ruserscan now develop performance code using taking advantage ofthe speed af Cr. (One Ruser was able to achieve aporirmance speed up ever 100K using Repp. Ifyou have advanced knontedge of Rand can think about production level code, youll Immediately make yourself more atracive to potential employers. [Acton Step: Werk through the Repp case study on R vectoration vs C++ vctorsation inthe Repp section. ‘Modify the function and try some new ones Take your facings and write ther up in an explanatory pos for a portolo prot 5. Introduction to Statistical Learning By: Gareth ames, Daniela Witen, Trevor Haste, and Robert Tissirant RUN ete Statistical CT Introduction te Statistical Learning sone of the best Introductory textbooks for machine learning. It provides easy o understand expanatons of concepts and coding examples th R Iealso covers the bass of linear modes extensively Iesimportant wo know these bases because these ae some ofthe mast common models Linear models are also popular in busines settings where medel interpretability s important. The effect that TV vs onine ad spendinghas on sales a perfect application of linear madels for erorexatity, Some other adevona topes covered include: + Kole cross-validation + Regularization + Feature selection + Polynomial regression + Tree based methods + Supportvector machines Unsupervised learning Action step: Use chapter on Clasiication o implementa logistic regression mode Use ths cra card eater to predict defaus. Tiss atypical appkation for data sles who work in sk management 6. The Elements of Statistical Learning By: Trevor Hastie, Robert shin Jerome Friedman Ifyou want te accelerate your machine learning career, you need to have a strong grasp on both fundamentals, and advanced topics The Elements of tatstial Learring i the perfect resource for bringing your machine learning skis tothe next eve Thists one of the most comprehensive books on machine earning This book reviews everyting fom linear metheds ta neural nets, boosting and random forests lesa bit moce mathy than other books, which is reat for galing a deeper understanding of the topics, [Dont ny to absorb the entire book st once though, Instead ake tn small hus, Pick topic na chaps, and bulla smal project (dort spend more than 8 10 hours. [Acton Step: Read Secon 343 and understand the diference between Ridge Regression and the asso. Use this housing tase to precet housing prices Use the SckI-Learn implementation of linear regression withallof the features, and then use Piso Rowrosson and the Lasso t9 Selec the most important features. 7. Understanding Machine Learning: From Theory to Algorithms '8y: Sha shalevshwarz and sha Ben Dawa Ifyou want a deeper understanding of machine learning algorithms, tis great book les splitinto the flowing sections of increasing comple + Foundations + From theory to algorithms + Aadeonal learning medlls + hovanced theory -Agreat way to gain a deep, lasting understanding of machine earring opiesis to np"? them from sera This the perfect reference fr implementing algorithms yourself Ifyou havent used a machine leaning model before, | dont recommend implementing From scratch righ aay. Star by using sckitearn or ne of Haris, and then after youve got ahandé on tty \weting 1 youset rom scratch This book prowiesexcersve theory onthe algortyms help you. ‘Acton sep: ead trcugh chapter 182 he decison re ago, hen fallow along waht onc re teal we your om for sth 8. Mining of Massive Datasets Bye Leche. Anand Rearama, ef Ukman PS \s Tits a grea book developed from various Stanford courses on large scale data miningand network ana, ‘The fous son datamining very large datasets Tiss important for implementing production level models at scale Large companies ke Google receive hundreds of millions (or more search queries per day so they ae especialy interested in mining very large datasets. Some topics covered inthis book indus * Mapreduce + Mining daca seams + Unkanalyss + Recommendation systems «Mining soiaLnetwork graphs + Dieensionality reduction + Large-scale machine learning Action Step: Read through chapterS on Link Anais. Theres a great example of haw Google uses the PageRank algorthm to assign a ea number toa pageto determinehow irmportant ei Compete exercise 5.1.1 te determine the PageRank of exch page nthe spin internet ‘modelin Figure 7. Use Python and ny to complete this exerlge, Dont forget to wet up as a portlio project. 9. Deep Learning By: an Goodretow, Yoshus Bengo, and Aaron Course Deep Ising sone ofthe hotest in machin esting. Companies ike Google, Facebook and Amazon need highly skied professionals with upertise in deep leaning, What eit that makes deep lesring 0 powerful? Irautomates one ofthe most fut parts of machine learning, feature discovery Rather than spendinghours ofime manually engineering new featuresin creative ways, deep learning automates he proces, youre new to deep learning ths book sa must. Even fyouhave some experience, those advanced deep leaning practitioners wl benefit as wo This books presence in an easyto read side format wth lat of bles and pues Here ae some ofthe topics covered + tno and explanation ofthe anporance of dep learning + Algorithms - backpropagation, conmmets, recurrent neural nets + Uncupenvees deep learning + acenton mechanisms [Action Step: Rea hrough the sector on agorthms and then use Pythors Vhs oy Alasiy WRIST digs using a multayer perceptron 10. Think Stats By: Allen 8. Downey Asa datascientisitsimportant that you have a solid grasp on probably and statistics. Machine learning model are rooted in the fundamentals of probability theory. Youll frequently be asked basi probabity and stats questions during interviews soit doesit hurtto refresh youre from time tte. Tis books geared towards programmes, so takes moce ofan applied approach rather than conventional textbooks that focus on the math and theory. Sections ae short ar easy to read so youl be ableto quik work through examples. Some ofthe topes covered include: Deserptive tatstes| + Cumulative detibuton functions + Cominuous distoutions Probailty + operation and distributions + Hypothesis testing Estimation + correlation [Acton Step: Read though chapter 7on hypothests testing. This chapter proves a good comparison between classical hypothesis testing and Bayesian hypathesi testing. ‘ork through exercise 73 o determine the posterior probably thatthe dstibution of th weight Is cfferent fort babies and others. Youll be working with data frm the sons) Survey ofFamy Grea WS 11. Bayesian Methods for Hackers By: Cam Davidson ion BAYESIAN METHODS Hackers ‘This a Bayesan Statist textbook that takes an “understanding fest’ "mathematle second” point of view ‘Bayesian inference fs an important rope in machine leaning that takes a diferent approach than classi inferetial stats, Tha Bayesian approach allows us to make inferences about things based on what we already know We can never be certain about an outcome, but with some pror knowledge, we can establish some confidence about an outcome, In arealworld seting, Bayesian statistic tapped to classification problems such a mall ficering spam” oF “not spam” and arte clasiicavon Ctecinolegy, sports, “pli Tiss an easy to read book, with frequent examples in Python code, The book has a ‘conversational tne, wich keeps things interesting Some toler neue: Bayesian methods + Moding Sayeian problems using Pyehon + Maciew Chain Monte Calo 1 Thelaw of large numbers + Loss functions + choosing appropriate prior estribatons [Acton Step: Reacthrough the example in Chapter on ayesian VB testing This ea great example ofa real-world application, NB testing i especally popularin online marketing does version A of awebsite get more sales than version ofthe website?) Code tis youselin Python, and play around with the number of was Nt see how the posterior dstrbution changes, 12. Think Bayes - Bayesian Statistics Made Simple By: Allen 8. Downey Mantels Bes Another great rasoure from Allen Dewney and Green Tes Press. This book takes a logical approach to sohing problems The author uses numerous examples to show you the types of decsions yout need to make when modling reat nord problem. Here ae some ofthe topes nus in this book + BayessTheorern + Computations eater # Decision analysis + Observer bias + Hypothesis testing + Dealing with dimensions [Action Step: Work through the NHL example in chapter 7 Code alongwith the examples n Python to compute the probably that each team win the ext game, Conclusion Now that you have alls ofthe best free data sclence resources to use in 2078, star bulding pupa Use the examples ve prowded to kickstart your earning Leave a comment tolet me know which projet youre going to build first Learnings a proces, sotate tin smal chunks. Dont pnd mere than TO hour larg ne slang buldng a project Remember, it doesnt hae tobe perfec The goal isto learn, bul and showcase your skis Fora quick review af althe project ideas, download the FREE quickstart projet guide below. ‘Bonus: Download a fee summary chest wth all 12 textbooks ta tare tearing immediately.

You might also like