book

Python: Advanced Predictive Analytics

by Ashish Kumar, Joseph Babcock

December 2017

Beginner to intermediate

660 pages

15h 31m

English

Packt Publishing

Read now

Unlock full access

What this learning path covers
Downloading the example codeErrataPiracyQuestions

Introducing predictive modellingScope of predictive modellingEnsemble of statistical algorithmsStatistical toolsHistorical dataMathematical functionBusiness contextKnowledge matrix for predictive modellingTask matrix for predictive modelling
LinkedIn's "People also viewed" featureWhat it does?How is it done?Correct targeting of online adsHow is it done?Santa Cruz predictive policingHow is it done?Determining the activity of a smartphone user using accelerometer dataHow is it done?Sport and fantasy leaguesHow was it done?
AnacondaStandalone PythonInstalling a Python packageInstalling pipInstalling Python packages with pip
Reading the data – variations and examplesData framesDelimiters
Case 1 – reading a dataset using the read_csv methodThe read_csv methodUse cases of the read_csv methodPassing the directory address and filename as variablesReading a .txt dataset with a comma delimiterSpecifying the column names of a dataset from a listCase 2 – reading a dataset using the open method of PythonReading a dataset line by lineChanging the delimiter of a datasetCase 3 – reading data from a URLCase 4 – miscellaneous casesReading from an .xls or .xlsx fileWriting to a CSV or Excel file
Checking for missing valuesWhat constitutes missing data?How missing values are generated and propagatedTreating missing valuesDeletionImputation
Scatter plotsHistogramsBoxplots
Subsetting a datasetSelecting columnsSelecting rowsSelecting a combination of rows and columnsCreating new columns
Various methods for generating random numbersSeeding a random numberGenerating random numbers following probability distributionsProbability density functionCumulative density functionUniform distributionNormal distributionUsing the Monte-Carlo simulation to find the value of piGeometry and mathematics behind the calculation of piGenerating a dummy data frame
AggregationFilteringTransformationMiscellaneous operations
Method 1 – using the Customer Churn ModelMethod 2 – using sklearnMethod 3 – using the shuffle function
Inner JoinLeft JoinRight JoinAn example of the Inner JoinAn example of the Left JoinAn example of the Right JoinSummary of Joins in terms of their length
Random sampling and the central limit theorem
Null versus alternate hypothesisZ-statistic and t-statisticConfidence intervals, significance levels, and p-valuesDifferent kinds of hypothesis testA step-by-step guide to do a hypothesis testAn example of a hypothesis test
Understanding the maths behind linear regressionLinear regression using simulated dataFitting a linear regression model and checking its efficacyFinding the optimum value of variable coefficients
p-valuesF-statisticsResidual Standard Error
Linear regression using the statsmodel libraryMultiple linear regressionMulti-collinearityVariance Inflation Factor
Training and testing data splitSummary of modelsLinear regression with scikit-learnFeature selection with scikit-learn
Handling categorical variablesTransforming a variable to fit non-linear relationsHandling outliersOther considerations and assumptions for linear regression
Linear regression versus logistic regression
Contingency tablesConditional probabilityOdds ratioMoving on to logistic regression from linear regressionEstimation using the Maximum Likelihood MethodLikelihood function:Log likelihood function:Building the logistic regression model from scratchMaking sense of logistic regression parametersWald testLikelihood Ratio Test statisticChi-square test
Processing the dataData explorationData visualizationCreating dummy variables for categorical variablesFeature selectionImplementing the model
Cross validation
The ROC curveConfusion matrix
Introduction to clustering – what, why, and how?What is clustering?How is clustering used?Why do we do clustering?
Distances between two observationsEuclidean distanceManhattan distanceMinkowski distanceThe distance matrixNormalizing the distancesLinkage methodsSingle linkageCompete linkageAverage linkageCentroid linkageWard's methodHierarchical clusteringK-means clustering
Importing and exploring the datasetNormalizing the values in the datasetHierarchical clustering using scikit-learnK-Means clustering using scikit-learnInterpreting the cluster
The elbow methodSilhouette Coefficient
Introducing decision treesA decision tree
HomogeneityEntropyInformation gainID3 algorithm to create a decision treeGini indexReduction in VariancePruning a treeHandling a continuous numerical variableHandling a missing value of an attribute
Visualizing the treeCross-validating and pruning the decision tree
Regression tree algorithmImplementing a regression tree using Python
The random forest algorithmImplementing a random forest using PythonWhy do random forests work?Important parameters for random forests
Best practices for codingCommenting the codesDefining functions for substantial individual tasksExample 1Example 2Example 3Avoid hard-coding of variables as much as possibleVersion controlUsing standard libraries, methods, and formulas
Designing an advanced analytic solutionData layer: warehouses, lakes, and streamsModeling layerDeployment layerReporting layer
Data input and transformationSanity checkingModel developmentScoringVisualization and reporting
Data input and transformationSanity checkingModel developmentScoringVisualization and reporting
Exploring categorical and numerical data in IPythonInstalling IPython notebookThe notebook interfaceLoading and inspecting dataBasic manipulations – grouping, filtering, mapping, and pivotingCharting with Matplotlib
Cleaning and convertingTime series diagnosticsJoining signals and correlation
Loading geospatial dataWorking in the cloud
Creating the SparkContextCreating an RDDCreating a Spark DataFrame
Similarity and distance metricsNumerical distance metricsCorrelation similarity metrics and time seriesSimilarity metrics for categorical dataK-means clustering
Where agglomerative clustering fails
Linear regressionData preparationModel fitting and evaluationStatistical significance of regression outputsGeneralize estimating equationsMixed effects modelsTime series dataGeneralized linear modelsApplying regularization to linear models
Decision treesRandom forest
Logistic regressionMulticlass logistic classifiers: multinomial regressionFormatting a dataset for classification problemsLearning pointwise updates with stochastic gradient descentJointly optimizing all parameters with second-order methods
Strategies for improving classification models
Fitting and SVM to the census dataBoosting – combining small models to improve accuracyGradient boosted decision trees
Working with textual dataCleaning textual dataExtracting features from textual dataUsing dimensionality reduction to simplify datasets
Latent Dirichlet AllocationUsing dimensionality reduction in predictive modeling
Cleaning image dataThresholding images to highlight objectsDimensionality reduction for image analysis
Learning patterns with neural networksA network of one – the perceptronCombining perceptrons – a single-layer neural networkParameter fitting with back-propagationDiscriminative versus generative modelsVanishing gradients and explaining awayPretraining belief networksUsing dropout to regularize networksConvolutional networks and rectified unitsCompressing Data with autoencoder networksOptimizing the learning rate
The MNIST dataConstructing the network
The architecture of a prediction service
The GET requestsThe POST requestThe HEAD requestThe PUT requestThe DELETE request
Application – the engine of the predictive services
Setting up the databaseThe web serverThe web applicationThe flow of a prediction service – training a modelOn-demand and bulk prediction
Checking the health of models with diagnosticsEvaluating changes in model performanceChanges in feature importanceChanges in unsupervised model performance
Experimental allocation – assigning customers to experimentsDeciding a sample sizeMultiple hypothesis testing
Translate terms to business valuesVisualizing resultsCase Study: building a reporting serviceThe report serverThe report applicationThe visualization layer

Content preview from Python: Advanced Predictive Analytics

Summary

In this chapter, we skimmed through the basic concepts of statistics. Here is a brief summary of the concepts we learned:

Hypothesis testing is used to test the statistical significance of a hypothesis. The one which already exists or is assumed to be true is a null hypothesis, the one which someone is not sure about or is being proposed as an alternate premise is an alternate hypothesis.
One needs to calculate a statistic and the associated p-value to conduct the test.
Hypothesis testing (p-values) is used to test the significance of the estimates of the coefficients calculated by the model.
The chi-square test is used to test the causal relationship between a predictor and an input variable. It can also be used to check whether the data ...

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Start your free trial

Mastering Predictive Analytics with Python

Joseph Babcock

Learning Predictive Analytics with Python

Ashish Kumar

Hands-On Predictive Analytics with Python

Alvaro Fuentes

Modeling Techniques in Predictive Analytics with Python and R: A Guide to Data Science

Thomas W. Miller

Publisher Resources

ISBN: 9781788992367Supplemental Content