Azure Machine Learning

DIG DEEP WITH AZURE MACHINE LEARNING Use data analysis to take your business to a whole new level. Microsoft Azure Machine Learning simplifies data analysis and empowers you to find the answers your business needs. The question isn't whether you can find the answers. The question is how. So, what do you want to find out? I WANT TO: Estimate product demand Regression Forecast the future by estimating the relationship between variables. Predict Values Predict sales figures Analyze marketing returns Predict credit risk Find Unusual Occurrences Anomaly Detection Identify and predict rare or unusual data points. Detect fraud Catch abnormal equipment readings (Do Perform customer segmentation Clustering Separate similar data points into intuitive groups. Discover Predict customer tastes Structure Determine market price Classification Identify what category new information belongs in. Predict Categories Predict Between Predict Between Two Categories Several Categories Two-Class Classification Multi-Class Classification Answers simple two-choice questions, like Answers complex questions with multiple possible answers. yes-or-no, true-or-false. Is this tweet positive? What is the mood of this tweet? Will this customer renew their service? Which service will this customer choose? Which of several promotions draws more customers? Which of two coupons A/B A/B/C draws more customers? Azure Machine Learning works by teaching the software to find patterns in the current data so that it can seek out the patterns in future data. Let's say you rent cars. How can you accurately predict demand for your product? FOR THAT YOU NEED REGRESSION ANALYSIS STEP 01 STEP 05 Predict future demand. Use the model to forecast Get the data. Car rental could spike depending on time of day, holidays, weather, etc. future spikes and shortfalls in demand. STEP 02 STEP 04 Prepare the data. Clean data, combine datasets, and prepare it for analysis. Score and evaluate the model. Test the model's ability to predict the original data, and evaluate its success. STEP 03 Train the model. Feed the information into the machine to teach it what to expect. Find out how to do this and more with #AzureML. Visit us at https://studio.azureml.net/ ALGORITHM MODULE OPTIONS Regression Ordinal Regression Poisson Regression Predicts event Fast forest quantile regression Predicts a distribution Linear Regression Fast training, linear model Data in rank ordered categories counts Example > Example > Example > Example > Y|X Bayesian Linear Regression Linear model, small Neural Network Decision Forest Boosted Decision Regression Accurate, long training times Regression Accurate, fast training times Tree Regression Accurate, fast training times, large memory footprint data sets Example > Example > Example > Example > Anomaly Detection One Class SVM PCA-Based Anomaly Detection Fast training times Under 100 features, aggressive boundary Example > Example > Clustering K-Means Unsupervised learning Example > Two-Class Classification Two-class averaged perceptron Fast training, linear model Two-class SVM Two-class logistic regression Fast training, linear model Two-class Bayes point machine Fast training, linear model Under 100 features, linear model Example > Example > Example > Example > Two-class decision forest Two-class boosted decision tree Two-class logistic regression Fast training, linear model Accurate, fast training Accurate, fast training, large memory footprint Example > Example > Example > Two-class Two-class Two-class decision jungle Accurate, small memory footprint locally deep SVM Under 100 neural network Accurate, long training times features Example > Example > Example > Multiclass Classification Multiclass logistic regression Fast training times, linear model Multiclass neural Multiclass network decision forest Accuracy, long training times Accuracy, fast training times Example > Example > Example > Multiclass One-v-all decision jungle Accuracy, small memory footprint multiclass Depends on the two-class classifier Example > Example >