Predicting the survival of a given passenger.
The sinking of the RMS Titanic is one of the most infamous shipwrecks in history. On April 15, 1912, during her maiden voyage, the Titanic sank after colliding with an iceberg, killing 1502 out of 2224 passengers and crew. This sensational tragedy shocked the international community and led to better safety regulations for ships. One of the reasons that the shipwreck led to such loss of life was that there were not enough lifeboats for the passengers and crew. Although there was some element of luck involved in surviving the sinking, some groups of people were more likely to survive than others, such as women, children, and the upper-class. In this challenge, we ask you to complete the analysis of what sorts of people were likely to survive. In particular, we ask you to apply the tools of machine learning to predict which passengers survived the tragedy. ![enter image description here] #What to do; - Create experiment - Create Classification model - Using Azure ML. - Using the Titanic passenger data set - Build a model for predicting the survival of a given passenger. #ML model when finished #AML model development step - Create: data preparation - Train: teach the algorithm with data - Score: see the performance - Evaluate: compare performance of each algorithm - Publish Web Service: production and re-train #Download Data set kaggle / Titanic: Machine Learning from Disaster https://www.kaggle.com/c/titanic/data #Data Dictionary Variable Definition Key survival Survival 0 = No, 1 = Yes pclass Ticket class 1 = 1st, 2 = 2nd, 3 = 3rd sex Sex Age Age in years sibsp # of siblings / spouses aboard the Titanic parch # of parents / children aboard the Titanic ticket Ticket number fare Passenger fare cabin Cabin number embarked Port of Embarkation C = Cherbourg, Q = Queenstown, S = Southampton #Variable Notes pclass: A proxy for socio-economic status (SES) - 1st = Upper - 2nd = Middle - 3rd = Lower age: Age is fractional if less than 1. If the age is estimated, is it in the form of xx.5 sibsp: The dataset defines family relations in this way... - Sibling = brother, sister, stepbrother, stepsister - Spouse = husband, wife (mistresses and fiancés were ignored) parch: The dataset defines family relations in this way... - Parent = mother, father - Child = daughter, son, stepdaughter, stepson - Some children travelled only with a nanny, therefore parch=0 for them. #View data set in Microsoft Excel #Machine Learning experiment creation working steps #Working steps; - Import Data set - Create New Experiment - Prepare Data - Drop the columns PassengerID, Name, Ticket, Cabin - Make categorical values: Survived, Pclass, Sex, Embarked - Replace missing value with median - Drop rows with missing data - Create Label - Split data 70% training and 30% scoring - Select Algorithm : Two-Class Boosted Decision - Train - Score #Import Data set 1. Click DATASETS 2. Click NEW 3. Click FROM LOCAL FILE #Upload a new dataset 4. Click Choose File 5. Brows and select train.csv 6. ENTER A NAME FOR THE NEW DATASET TitanicTrain1 7. SELECT A TYPE FOR THE NEW DATASET Generic CSV File with a header (.csv) 8. PROVIDE AN OPTIONAL DESCRIPTION kaggle Titanic: Machine Leering from disaster 9. Click #Verify dataset uploaded 1. Click DATASETS 2. Make sure TitanicTrain1 is in MY DATASETS list #Create New Experiment #Create Blank Experiment #Set experiment name Type in name = Titanic 1 #Drag & drop dataset to canvas 1. Click Saved Datasets / My Datasets 2. Drag & drop TitanicTrain1 to canvas 3. Visualize output Drop the columns PassengerID, Name, Ticket, Cabin 1. Drag & drop module Select Columns in Dataset 2. Selected columne = Drop Columns: PassengerId, Name, Cabin, Ticket 3. Click Launch column selector 4. Visualize #Drop the columns PassengerID, Name, Ticket, Cabin 5. Begin With = ALL COLUMNS / Exclude / column name 6. Selected column PassengerID, Name, Ticket, Cabin 7. Click 8. Visualize #Make categorical values: Survived, Pclass, Sex, Embarked 1. Drag & drop Edit Metadata 2. Comment = Cast and Rename Columns: Survived, PClass, Sex, Embarked 3. Selected column Survived, Pclass, Sex, Embarked 4. Data type = Unchanged 5. Categorical = Make categorical 6. Fields = Unchanged 7. New column name = Survived, PassengerClass, Gender, PortEmbarkation 8. Visualize #Rename columns 1. Drag & drop Edit Metadata 2. Comment = Rename Columns: SiblingSpouse, ParentChild, FarePrice 3. Selected column SibSp, Parch, Fare 4. Data type = Unchanged 5. Categorical = Unchanged 6. Fields = Unchanged 7. New column name = SiblingSpouse, ParentChild, FarePrice 8. Visualize #Replace missing value with median 1. Drag & drop Missing Values Scrubber 2. Comment = Replace missing value with median 3. Set properties 4. Visualize #Drop rows with missing data 1. Drag & drop Missing Values Scrubber 2. Comment = Drop rows with missing values 3. Set properties 4. Visualize #Create Label 1. Drag & drop Edit Metadata 2. Comment = Assigning target variable 3. Selected column = Survived 4. Data type = Unchanged 5. Categorical = Unchanged 6. Fields = Label 7. New column name = - 8. Visualize #Import and Dataset preparation #Split data 70% training and 30% scoring 1. Drag & drop Split data 2. Set properties #Add Algorithm, Train and Score - Add Two-Class Boosted Decision tree - Add Train Model - Add Score Model #Add Two-Class Boosted Decision tree 1. Drag & drop Two-Class Boosted Decision tree 2. Set properties #Add Train Model 1. Drag & drop train model 2. Set column to Survived #Add Score Mode 1. Drag & drop Score Model 2. Set property = Append score column 3. Save 4. Run experiment 5. Visualize #Create web service 1. Save as Titanic 2 2. Run Titanic 2 3. Click SET UP WEB SERVICE 4. Click Predictive Web Service 5. Click RUN 6. Click SET UP WEB SERVICE #Create web service Titanic 2 [predictive exp.] page #Test web service #Web service testing - REQUEST/RESPONSE Test - REQUEST/RESPONSE Test preview - REQUEST/RESPONSE Excel workbook test - BATCH EXECUTION Test preview - BATCH EXECUTION Excel workbook test #REQUEST/RESPONSE Test 1. Test with know result 2. Open file kaggle test.csv 3. Take one passenger 4. Click REQUEST/RESPONSE Test 5. Fill in the form #REQUEST/RESPONSE Test preview 1. Test with know result 2. Open file kaggle test.csv 3. Take one passenger 4. Click REQUEST/RESPONSE Test preview 5. Click Enable (Sample Data) 6. Fill in the form 7. Click Test Request-Response #REQUEST/RESPONSE Excel workbook test 1. Test with know result 2. Open file kaggle test.csv 3. Take one passenger 4. Click REQUEST/RESPONSE Excel 2013 or later 5. Open file Titanic 2 [Predictive Exp.] on Desktop 6. Click Enable Editing #REQUEST/RESPONSE Excel workbook test 1. Input = Sheet1!A4L4 2. My data has headers = uncheck 3. Output = Sheet1!A1 4. Include headers = check 5. Copy a line from file kaggle test.csv to A4 6. Click Predict #REQUEST/RESPONSE Excel workbook test #Test result #More information **More information on Classification Model** Two-Class Boosted Decision Tree https://msdn.microsoft.com/en-us/library/azure/dn906025.aspx Machine learning algorithm cheat sheet for Microsoft Azure Machine Learning Studio https://docs.microsoft.com/en-us/azure/machine-learning/machine-learning-algorithm-cheat-sheet <br><br> ---------- > This ML experiment is for [Microsoft Azure Machine Learning Course].<br> For the complete experiment list [Click here].<br> Laploy | firstname.lastname@example.org | 084 007 5544 | [www.laploy.com]<br> ![enter image description here] ---------- : https://notebooks.azure.com/laploy/libraries/loyml/html/00001%20Sessions%20summary.ipynb : https://gallery.cortanaintelligence.com/Home/Author?authorId=81E333F747E3429B55A3445E6714C36F60B397C13B4D0B07F34DEF1421F64D73 : http://laploy.com : https://raw.githubusercontent.com/laploy/mli/master//loy-small.jpg : https://raw.githubusercontent.com/laploy/mli/master/10600-001.JPG : https://raw.githubusercontent.com/laploy/mli/master//loy-small.jpg