SlideShare a Scribd company logo
Topic: Churn Prediction for Sleep Apnea
study.
--Rahul Bhojwani (bhojw005@umn.edu)
Data Definition
● Daily device usage time
● Daily Avg Pressure readings from the device
● Daily Apnea Hypopnea index
● Daily device leakage readings
● Patients demographic data
● Diagnosis test results from timely hospital checkups
Problem Definition:
● Sleep Apnea Study at Fairview Hospital.
● ~22 million Americans suffer from sleep apnea.
● More than 200,000 US cases registered per year
● CPAP device prescribed
● Patient should stick for at least 6 months
● In General adherence is ~40-50%. Hospital resources
are making calls helping it increase to ~65%.
GOAL:
Predict as early as possible whether a patient will adhere to the
CPAP device usage. Churn is defined based on the usage in
151-180th day.
● Model design:
○ 30 days -> 180 days (Habit Problem)
○ 13 days -> 180 days
● Data exploration and Feature extraction:
○ Raw time series
○ Mean and SD over intervals
○ Change in mean / slope
○ Demographic data
○ % days adherence
● Imbalanced class problem:
○ Undersampling
○ Oversampling- SMOTE
○ Balanced Cascade
● Feature Selection:
○ Anova
○ Chi-square
○ L1
○ RFE
● Models experimented with, on structured
data:
○ Decision Tree
○ SVM
○ Random Forest
○ XGBoost
○ Deep learning Feed forward Network
● Optimized for:
○ F1 score
○ npv, ppv
○ Cohen Kappa
● Time Series Classification:
○ 3 daily time series of:
■ Usage time
■ Device Pressure
■ AHI events
○ Demographic and diagnosis data
● Time series similarity:
○ Cosine similarity with average time series of each
class. ( Problem coz of varying time series)
○ Smoothed time series cosine similarity.
■ Expanding mean
● Deep Learning models:
○ LSTM for time series classification
○ CNN
○ Long Short Term Memory-Fully Convolutional
Neural Network (LSTM-FCNN)
(https://arxiv.org/abs/1709.05206)
○ Attention Long Short Term Memory-Fully
Convolutional Neural Network (ALSTM-FCNN)
(https://arxiv.org/abs/1709.05206)
● Conclusion:-
○ Using the data of first 13 days, we can predict the
engagement condition after 6 months with a decent
accuracy
○ In initial 13 days of CPAP prescription, decision to
continue with the treatment or suggesting an alternative
one can be taken.

More Related Content

Similar to Churn prediction for a sleep apnea study

ReComp and P4@NU: Reproducible Data Science for Health
ReComp and P4@NU: Reproducible Data Science for HealthReComp and P4@NU: Reproducible Data Science for Health
ReComp and P4@NU: Reproducible Data Science for Health
Paolo Missier
 
In-memory Applications for Oncology
In-memory Applications for OncologyIn-memory Applications for Oncology
In-memory Applications for OncologyMatthieu Schapranow
 
Alz Hack II
Alz Hack IIAlz Hack II
Alz Hack II
Frank Kelly
 
PubChem for drug discovery in the age of big data and artificial intelligence
PubChem for drug discovery in the age of big data and artificial intelligencePubChem for drug discovery in the age of big data and artificial intelligence
PubChem for drug discovery in the age of big data and artificial intelligence
Sunghwan Kim
 
Machine learning and Internet of Things, the future of medical prevention
Machine learning and Internet of Things, the future of medical preventionMachine learning and Internet of Things, the future of medical prevention
Machine learning and Internet of Things, the future of medical prevention
Pierre Gutierrez
 
Progressive Transformer-Based Generation of Radiology Reports
Progressive Transformer-Based Generation of Radiology ReportsProgressive Transformer-Based Generation of Radiology Reports
Progressive Transformer-Based Generation of Radiology Reports
KristophesCowie
 
Enhancing the Prediction Accuracy of Solar Power Generation using a Generativ...
Enhancing the Prediction Accuracy of Solar Power Generation using a Generativ...Enhancing the Prediction Accuracy of Solar Power Generation using a Generativ...
Enhancing the Prediction Accuracy of Solar Power Generation using a Generativ...
Kundjanasith Thonglek
 
Genomic Big Data Management, Integration and Mining - Emanuel Weitschek
Genomic Big Data Management, Integration and Mining - Emanuel WeitschekGenomic Big Data Management, Integration and Mining - Emanuel Weitschek
Genomic Big Data Management, Integration and Mining - Emanuel Weitschek
Data Driven Innovation
 
Appendicectomy national meeting
Appendicectomy national meetingAppendicectomy national meeting
Appendicectomy national meetinganeelbhangu
 
Experimental Design Considerations to Optimize Chronic Cardiovascular Telemet...
Experimental Design Considerations to Optimize Chronic Cardiovascular Telemet...Experimental Design Considerations to Optimize Chronic Cardiovascular Telemet...
Experimental Design Considerations to Optimize Chronic Cardiovascular Telemet...
InsideScientific
 
Electricity Consumption Forecasting Using Arima, UCM, Machine Learning and De...
Electricity Consumption Forecasting Using Arima, UCM, Machine Learning and De...Electricity Consumption Forecasting Using Arima, UCM, Machine Learning and De...
Electricity Consumption Forecasting Using Arima, UCM, Machine Learning and De...
Giorgio Carbone
 
Artificial Intelligence Chapter 9 Negnevitsky
Artificial Intelligence Chapter 9 NegnevitskyArtificial Intelligence Chapter 9 Negnevitsky
Artificial Intelligence Chapter 9 Negnevitskylopanath
 
Oxford Lectures Part 1
Oxford Lectures Part 1Oxford Lectures Part 1
Oxford Lectures Part 1Andrea Pasqua
 
Make Sense Out of Data with Feature Engineering
Make Sense Out of Data with Feature EngineeringMake Sense Out of Data with Feature Engineering
Make Sense Out of Data with Feature Engineering
DataRobot
 
Endobronchial Ultrasound Image Diagnosis Using Convolutional Neural Network (...
Endobronchial Ultrasound Image Diagnosis Using Convolutional Neural Network (...Endobronchial Ultrasound Image Diagnosis Using Convolutional Neural Network (...
Endobronchial Ultrasound Image Diagnosis Using Convolutional Neural Network (...
Yan-Wei Lee
 
Ppt dawn trial
Ppt dawn trialPpt dawn trial
Ppt dawn trial
NeurologyKota
 
Cri big data
Cri big dataCri big data
Cri big data
Putchong Uthayopas
 
OSFair2017 Training | Reproducibility in critical care research
OSFair2017 Training | Reproducibility in critical care researchOSFair2017 Training | Reproducibility in critical care research
OSFair2017 Training | Reproducibility in critical care research
Open Science Fair
 
ELRIG Event Biocity Scotland May19
ELRIG Event Biocity Scotland May19ELRIG Event Biocity Scotland May19
ELRIG Event Biocity Scotland May19
Angelo Pugliese
 
Predicting Thyroid Disorder with Deep Neural Networks
Predicting Thyroid Disorder with Deep Neural NetworksPredicting Thyroid Disorder with Deep Neural Networks
Predicting Thyroid Disorder with Deep Neural Networks
Anaelia Ovalle
 

Similar to Churn prediction for a sleep apnea study (20)

ReComp and P4@NU: Reproducible Data Science for Health
ReComp and P4@NU: Reproducible Data Science for HealthReComp and P4@NU: Reproducible Data Science for Health
ReComp and P4@NU: Reproducible Data Science for Health
 
In-memory Applications for Oncology
In-memory Applications for OncologyIn-memory Applications for Oncology
In-memory Applications for Oncology
 
Alz Hack II
Alz Hack IIAlz Hack II
Alz Hack II
 
PubChem for drug discovery in the age of big data and artificial intelligence
PubChem for drug discovery in the age of big data and artificial intelligencePubChem for drug discovery in the age of big data and artificial intelligence
PubChem for drug discovery in the age of big data and artificial intelligence
 
Machine learning and Internet of Things, the future of medical prevention
Machine learning and Internet of Things, the future of medical preventionMachine learning and Internet of Things, the future of medical prevention
Machine learning and Internet of Things, the future of medical prevention
 
Progressive Transformer-Based Generation of Radiology Reports
Progressive Transformer-Based Generation of Radiology ReportsProgressive Transformer-Based Generation of Radiology Reports
Progressive Transformer-Based Generation of Radiology Reports
 
Enhancing the Prediction Accuracy of Solar Power Generation using a Generativ...
Enhancing the Prediction Accuracy of Solar Power Generation using a Generativ...Enhancing the Prediction Accuracy of Solar Power Generation using a Generativ...
Enhancing the Prediction Accuracy of Solar Power Generation using a Generativ...
 
Genomic Big Data Management, Integration and Mining - Emanuel Weitschek
Genomic Big Data Management, Integration and Mining - Emanuel WeitschekGenomic Big Data Management, Integration and Mining - Emanuel Weitschek
Genomic Big Data Management, Integration and Mining - Emanuel Weitschek
 
Appendicectomy national meeting
Appendicectomy national meetingAppendicectomy national meeting
Appendicectomy national meeting
 
Experimental Design Considerations to Optimize Chronic Cardiovascular Telemet...
Experimental Design Considerations to Optimize Chronic Cardiovascular Telemet...Experimental Design Considerations to Optimize Chronic Cardiovascular Telemet...
Experimental Design Considerations to Optimize Chronic Cardiovascular Telemet...
 
Electricity Consumption Forecasting Using Arima, UCM, Machine Learning and De...
Electricity Consumption Forecasting Using Arima, UCM, Machine Learning and De...Electricity Consumption Forecasting Using Arima, UCM, Machine Learning and De...
Electricity Consumption Forecasting Using Arima, UCM, Machine Learning and De...
 
Artificial Intelligence Chapter 9 Negnevitsky
Artificial Intelligence Chapter 9 NegnevitskyArtificial Intelligence Chapter 9 Negnevitsky
Artificial Intelligence Chapter 9 Negnevitsky
 
Oxford Lectures Part 1
Oxford Lectures Part 1Oxford Lectures Part 1
Oxford Lectures Part 1
 
Make Sense Out of Data with Feature Engineering
Make Sense Out of Data with Feature EngineeringMake Sense Out of Data with Feature Engineering
Make Sense Out of Data with Feature Engineering
 
Endobronchial Ultrasound Image Diagnosis Using Convolutional Neural Network (...
Endobronchial Ultrasound Image Diagnosis Using Convolutional Neural Network (...Endobronchial Ultrasound Image Diagnosis Using Convolutional Neural Network (...
Endobronchial Ultrasound Image Diagnosis Using Convolutional Neural Network (...
 
Ppt dawn trial
Ppt dawn trialPpt dawn trial
Ppt dawn trial
 
Cri big data
Cri big dataCri big data
Cri big data
 
OSFair2017 Training | Reproducibility in critical care research
OSFair2017 Training | Reproducibility in critical care researchOSFair2017 Training | Reproducibility in critical care research
OSFair2017 Training | Reproducibility in critical care research
 
ELRIG Event Biocity Scotland May19
ELRIG Event Biocity Scotland May19ELRIG Event Biocity Scotland May19
ELRIG Event Biocity Scotland May19
 
Predicting Thyroid Disorder with Deep Neural Networks
Predicting Thyroid Disorder with Deep Neural NetworksPredicting Thyroid Disorder with Deep Neural Networks
Predicting Thyroid Disorder with Deep Neural Networks
 

Recently uploaded

Everything you wanted to know about LIHTC
Everything you wanted to know about LIHTCEverything you wanted to know about LIHTC
Everything you wanted to know about LIHTC
Roger Valdez
 
一比一原版(Deakin毕业证书)迪肯大学毕业证如何办理
一比一原版(Deakin毕业证书)迪肯大学毕业证如何办理一比一原版(Deakin毕业证书)迪肯大学毕业证如何办理
一比一原版(Deakin毕业证书)迪肯大学毕业证如何办理
oz8q3jxlp
 
一比一原版(UofS毕业证书)萨省大学毕业证如何办理
一比一原版(UofS毕业证书)萨省大学毕业证如何办理一比一原版(UofS毕业证书)萨省大学毕业证如何办理
一比一原版(UofS毕业证书)萨省大学毕业证如何办理
v3tuleee
 
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
John Andrews
 
一比一原版(Bradford毕业证书)布拉德福德大学毕业证如何办理
一比一原版(Bradford毕业证书)布拉德福德大学毕业证如何办理一比一原版(Bradford毕业证书)布拉德福德大学毕业证如何办理
一比一原版(Bradford毕业证书)布拉德福德大学毕业证如何办理
mbawufebxi
 
Influence of Marketing Strategy and Market Competition on Business Plan
Influence of Marketing Strategy and Market Competition on Business PlanInfluence of Marketing Strategy and Market Competition on Business Plan
Influence of Marketing Strategy and Market Competition on Business Plan
jerlynmaetalle
 
原版制作(Deakin毕业证书)迪肯大学毕业证学位证一模一样
原版制作(Deakin毕业证书)迪肯大学毕业证学位证一模一样原版制作(Deakin毕业证书)迪肯大学毕业证学位证一模一样
原版制作(Deakin毕业证书)迪肯大学毕业证学位证一模一样
u86oixdj
 
Adjusting OpenMP PageRank : SHORT REPORT / NOTES
Adjusting OpenMP PageRank : SHORT REPORT / NOTESAdjusting OpenMP PageRank : SHORT REPORT / NOTES
Adjusting OpenMP PageRank : SHORT REPORT / NOTES
Subhajit Sahu
 
The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...
jerlynmaetalle
 
Criminal IP - Threat Hunting Webinar.pdf
Criminal IP - Threat Hunting Webinar.pdfCriminal IP - Threat Hunting Webinar.pdf
Criminal IP - Threat Hunting Webinar.pdf
Criminal IP
 
一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单
ewymefz
 
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
AbhimanyuSinha9
 
Algorithmic optimizations for Dynamic Levelwise PageRank (from STICD) : SHORT...
Algorithmic optimizations for Dynamic Levelwise PageRank (from STICD) : SHORT...Algorithmic optimizations for Dynamic Levelwise PageRank (from STICD) : SHORT...
Algorithmic optimizations for Dynamic Levelwise PageRank (from STICD) : SHORT...
Subhajit Sahu
 
Data_and_Analytics_Essentials_Architect_an_Analytics_Platform.pptx
Data_and_Analytics_Essentials_Architect_an_Analytics_Platform.pptxData_and_Analytics_Essentials_Architect_an_Analytics_Platform.pptx
Data_and_Analytics_Essentials_Architect_an_Analytics_Platform.pptx
AnirbanRoy608946
 
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
2023240532
 
Machine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptxMachine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptx
balafet
 
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
ahzuo
 
My burning issue is homelessness K.C.M.O.
My burning issue is homelessness K.C.M.O.My burning issue is homelessness K.C.M.O.
My burning issue is homelessness K.C.M.O.
rwarrenll
 
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Subhajit Sahu
 
哪里卖(usq毕业证书)南昆士兰大学毕业证研究生文凭证书托福证书原版一模一样
哪里卖(usq毕业证书)南昆士兰大学毕业证研究生文凭证书托福证书原版一模一样哪里卖(usq毕业证书)南昆士兰大学毕业证研究生文凭证书托福证书原版一模一样
哪里卖(usq毕业证书)南昆士兰大学毕业证研究生文凭证书托福证书原版一模一样
axoqas
 

Recently uploaded (20)

Everything you wanted to know about LIHTC
Everything you wanted to know about LIHTCEverything you wanted to know about LIHTC
Everything you wanted to know about LIHTC
 
一比一原版(Deakin毕业证书)迪肯大学毕业证如何办理
一比一原版(Deakin毕业证书)迪肯大学毕业证如何办理一比一原版(Deakin毕业证书)迪肯大学毕业证如何办理
一比一原版(Deakin毕业证书)迪肯大学毕业证如何办理
 
一比一原版(UofS毕业证书)萨省大学毕业证如何办理
一比一原版(UofS毕业证书)萨省大学毕业证如何办理一比一原版(UofS毕业证书)萨省大学毕业证如何办理
一比一原版(UofS毕业证书)萨省大学毕业证如何办理
 
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
 
一比一原版(Bradford毕业证书)布拉德福德大学毕业证如何办理
一比一原版(Bradford毕业证书)布拉德福德大学毕业证如何办理一比一原版(Bradford毕业证书)布拉德福德大学毕业证如何办理
一比一原版(Bradford毕业证书)布拉德福德大学毕业证如何办理
 
Influence of Marketing Strategy and Market Competition on Business Plan
Influence of Marketing Strategy and Market Competition on Business PlanInfluence of Marketing Strategy and Market Competition on Business Plan
Influence of Marketing Strategy and Market Competition on Business Plan
 
原版制作(Deakin毕业证书)迪肯大学毕业证学位证一模一样
原版制作(Deakin毕业证书)迪肯大学毕业证学位证一模一样原版制作(Deakin毕业证书)迪肯大学毕业证学位证一模一样
原版制作(Deakin毕业证书)迪肯大学毕业证学位证一模一样
 
Adjusting OpenMP PageRank : SHORT REPORT / NOTES
Adjusting OpenMP PageRank : SHORT REPORT / NOTESAdjusting OpenMP PageRank : SHORT REPORT / NOTES
Adjusting OpenMP PageRank : SHORT REPORT / NOTES
 
The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...
 
Criminal IP - Threat Hunting Webinar.pdf
Criminal IP - Threat Hunting Webinar.pdfCriminal IP - Threat Hunting Webinar.pdf
Criminal IP - Threat Hunting Webinar.pdf
 
一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单
 
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
 
Algorithmic optimizations for Dynamic Levelwise PageRank (from STICD) : SHORT...
Algorithmic optimizations for Dynamic Levelwise PageRank (from STICD) : SHORT...Algorithmic optimizations for Dynamic Levelwise PageRank (from STICD) : SHORT...
Algorithmic optimizations for Dynamic Levelwise PageRank (from STICD) : SHORT...
 
Data_and_Analytics_Essentials_Architect_an_Analytics_Platform.pptx
Data_and_Analytics_Essentials_Architect_an_Analytics_Platform.pptxData_and_Analytics_Essentials_Architect_an_Analytics_Platform.pptx
Data_and_Analytics_Essentials_Architect_an_Analytics_Platform.pptx
 
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
Quantitative Data AnalysisReliability Analysis (Cronbach Alpha) Common Method...
 
Machine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptxMachine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptx
 
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
一比一原版(CBU毕业证)卡普顿大学毕业证如何办理
 
My burning issue is homelessness K.C.M.O.
My burning issue is homelessness K.C.M.O.My burning issue is homelessness K.C.M.O.
My burning issue is homelessness K.C.M.O.
 
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
 
哪里卖(usq毕业证书)南昆士兰大学毕业证研究生文凭证书托福证书原版一模一样
哪里卖(usq毕业证书)南昆士兰大学毕业证研究生文凭证书托福证书原版一模一样哪里卖(usq毕业证书)南昆士兰大学毕业证研究生文凭证书托福证书原版一模一样
哪里卖(usq毕业证书)南昆士兰大学毕业证研究生文凭证书托福证书原版一模一样
 

Churn prediction for a sleep apnea study

  • 1. Topic: Churn Prediction for Sleep Apnea study. --Rahul Bhojwani (bhojw005@umn.edu) Data Definition ● Daily device usage time ● Daily Avg Pressure readings from the device ● Daily Apnea Hypopnea index ● Daily device leakage readings ● Patients demographic data ● Diagnosis test results from timely hospital checkups Problem Definition: ● Sleep Apnea Study at Fairview Hospital. ● ~22 million Americans suffer from sleep apnea. ● More than 200,000 US cases registered per year ● CPAP device prescribed ● Patient should stick for at least 6 months ● In General adherence is ~40-50%. Hospital resources are making calls helping it increase to ~65%. GOAL: Predict as early as possible whether a patient will adhere to the CPAP device usage. Churn is defined based on the usage in 151-180th day.
  • 2. ● Model design: ○ 30 days -> 180 days (Habit Problem) ○ 13 days -> 180 days ● Data exploration and Feature extraction: ○ Raw time series ○ Mean and SD over intervals ○ Change in mean / slope ○ Demographic data ○ % days adherence ● Imbalanced class problem: ○ Undersampling ○ Oversampling- SMOTE ○ Balanced Cascade ● Feature Selection: ○ Anova ○ Chi-square ○ L1 ○ RFE ● Models experimented with, on structured data: ○ Decision Tree ○ SVM ○ Random Forest ○ XGBoost ○ Deep learning Feed forward Network ● Optimized for: ○ F1 score ○ npv, ppv ○ Cohen Kappa
  • 3. ● Time Series Classification: ○ 3 daily time series of: ■ Usage time ■ Device Pressure ■ AHI events ○ Demographic and diagnosis data ● Time series similarity: ○ Cosine similarity with average time series of each class. ( Problem coz of varying time series) ○ Smoothed time series cosine similarity. ■ Expanding mean ● Deep Learning models: ○ LSTM for time series classification ○ CNN ○ Long Short Term Memory-Fully Convolutional Neural Network (LSTM-FCNN) (https://arxiv.org/abs/1709.05206) ○ Attention Long Short Term Memory-Fully Convolutional Neural Network (ALSTM-FCNN) (https://arxiv.org/abs/1709.05206) ● Conclusion:- ○ Using the data of first 13 days, we can predict the engagement condition after 6 months with a decent accuracy ○ In initial 13 days of CPAP prescription, decision to continue with the treatment or suggesting an alternative one can be taken.