Business Analytics Assignment (Solved)
12 Pages1793 Words236 Views
Added on 2020-12-29
Business Analytics Assignment (Solved)
Added on 2020-12-29
ShareRelated Documents
BUSINESS ANALYTICS
TABLE OF CONTENTSSECTION A.....................................................................................................................................21 Explaining how to measure accuracy of classification model.................................................22. Application of logistic regre...................................................................................................4SECTION B.....................................................................................................................................53.a Steps for developing regression model for predicting spending more than $700.................53.b Developing predictive model for predicting spend amount of a new male customer withspecified criteria..........................................................................................................................64.a Analysing data with reference to insights and recommendation...........................................64.b Other data could be added for making it more useful...........................................................65.a What predictive model is suggested related to risk of diabetes to age, weight and gender ofperson..........................................................................................................................................75.b Developing estimated regression model...............................................................................76.a..............................................................................................................................................106.b..............................................................................................................................................10REFERENCES..............................................................................................................................11
SECTION A1 Explaining how to measure accuracy of classification modelClassification model attempts for drawing conclusion through observed values as thiswill try for predicting value of one or more outcomes. Table 1Confusion matrixActualPredictedTrue PositiveFalse PositiveFalse NegativeTrue NegativeAccuracy(TP + TN)/ (TP + TN + FP + FN)PrecessionTP / (TP +FP)RecallTP / (TP + FN)True Positive rateTP / (TP + FN)False Positive rateFP/ (FP + Tn)It could be measured with key testing metrics which is stated below:Confusion matrix: It is referred as matrix representation of outcomes of binary testing(Confusion matrix, 2019). For instance, consider case of forecasting malaria so in this contextmedical testing has been done with outcome of some specific tests. In actual, one is going forvalidating hypothesis of declaring that person has malaria is acceptable of not. With theassumptions of 100 people, 20 people have malaria but in real scenario, only 15 has malaria andin those 15 also, 12 people were diagnosed correctly. Let, put this outcome in confusion matrix:Table 2Confusion matrixActualHaving malariaNot having malariaPredictedHaving malaria128Not having malaria377With combination of table 1 and 2,True positive: 12 (Correctly predicted)True negative: 77 (Correctly predicted)False positive: 8 (Low risk error)False negative: 3 (Type-I error)
With context of accuracy of prediction model, the outcome must be in ratio of accuratelypredicted number along with total number of people i.e. (12+77)/100 which is 0.89 with study ofconfusion matrix, it has been extracted:On basis of top row depicting the aggregate of prediction for having malaria. In thesepredictions, it had been forecasted correctly that 12 people are correctly having malaria inactual. So, ratio 12/ (12 + 8) is 0.6 is accuracy measure of model for detecting person tohave malaria which is known as precision of model.With consideration of first column, it signifies total number of people having malaria inactual and it has been forecasted that 12 are correct. Henceforth, ratio 12/ (12 + 3) is 0.8as accuracy measure of model for detecting person out of all people who have malaria inactual is termed as recall.
End of preview
Want to access all the pages? Upload your documents or become a member.
Related Documents
Evaluating Models for Customer Churn Predictionlg...
|12
|1656
|101
A confusion matrix is a table that is often used to describe thelg...
|2
|302
|301
Analytics: Questions with Answerslg...
|13
|1710
|92
Understanding Classifier Performance: Metrics and ROC Curvelg...
|4
|655
|369
Introduction to Clinical Epidemiology: Analysis of Diagnostic Test and Research Articlelg...
|12
|2139
|231
Weka: Data Mining and Machine Learning Topic 2022lg...
|13
|1214
|16