Predicting the Assisted Living Care Needs Using Machine Learning and Health State Survey Data

Predicting the Assisted Living Care Needs Using Machine Learning and Health State Survey Data

Table of Contents


Effective pain management can significantly improve the quality of life and outcomes for various types of patients(e.g. elderly, adult, young) and often requires assisted living for a significant number of people worldwide. In order to improve our understanding of patients’ responses to pain and needs for assisted living we need to develop adequate data processing techniques that would enable us to understand underlying interdependencies. To this purpose in this paper, we develop several different algorithms that can predict the need for medically assisted living outcomes using a large database obtained as a part of the national health survey. As a part of the survey, the respondents provided detailed information about general health care state, acute and chronic problems as well as the personal perception of pain associated with performing two simple talks: walking on the flat surface and walking upstairs. We model the correspondent responses using multinomial random variables and propose structured deep learning models based on maximum likelihood estimation and machine learning for information fusion. For comparison purposes, we also implement a fully connected deep learning network and use its results as benchmark measurements. We evaluate the performance of the proposed techniques using the national survey data and split them into two parts used training and testing. Our preliminary results indicate that the proposed models can potentially be useful in forecasting the need for medically assisted living. Clinical relevance— The proposed results can potentially be used to forecast the resources needed to offer long-term medical care to both chronic and acute conditions patients not being able to perform daily tasks independently.

  • IEEE Keywords

    • Machine learning,
    • Pain,
    • Assisted living,
    • Maximum likelihood estimation,
    • Accidents,
    • Diseases,
    • Detectors
  • Controlled Indexing

    • assisted living,
    • geriatrics,
    • health care,
    • learning (artificial intelligence),
    • maximum likelihood estimation,
    • neural nets


The issue of assisted living (AL) has been a subjectof significant public concern in both North America andworldwide. Consequently, analysis and forecasting of theassisted living needs have been subject of considerableresearch interest in the recent years [1]-[3]. The focus ofthe research has been quite extensive including forecastingthe financial cost to the public health care and insurancesystems, analyzing and forecasting type of services that arecommonly needed, analyzing social, financial and psycho-logical burden experienced by both long-term care patientsand their families. As discussed in the seminal paper relatedto the long-term care trend prediction [4] a potentially usefulapproach would consist of a longitudinal study rather than a commonly used single time-point approach and to thispurpose there were proposals of including admission cohortto the National Nursing Home Surveys [4]. However the lastNational Home Survey in the United States was done in 2004and hence may not represent adequately the current state ofhealth nor is in concordance with current design of healthsurveys.In general the analysis of long-term care needs is done us-ing surveys which include multi-level responses to questionswhich can be modelled as multinomial random variables.Consequently, commonly used technique in analyzing thesedata sets are based on multinomial logistic regression orany of the corresponding variants in the variety of thefields [7], [8], [9]. Most of these statistical analysis toolsbelong to a wide group of so called softmax regressiontools. These tools have been recently utilized extensively indeep learning neural networks as they are quite amenable tofeature extraction algorithms applied to highly dimensionaldata sets. To this purpose in this paper we develop deeplearning based networks for analyzing national health surveydata in order to predict the need for the medical assistancein performing the daily tasks.

In our previous work we developed an age-dependentmultinomial regression model to estimate the function relatedpain outcomes using the third national study in Serbia “Na-tional Health Status Survey in 2013” [6]. We demonstratedability to predict the pain related outcomes in the elderlygroup of patients that received medical help due to theinjuries obtained in several types of accidents. We performedthe data analysis on the subset consisting of 899 patients thatrequired medical attention immediately after the accident. Inthis paper we propose an unsupervised algorithm that can po-tentially be used for predicting medical care requirements ona general population including all the data points (subjects) inthe survey. To achieve this goal in the first part we proposethree different deep learning (DL) architectures: structuredDL network using unsupervised maximum likelihood fusion,structured DL network using supervised machine learning,and fully connected deep learning network using all theparameters. In Section 3 we present the results of theproposed algorithms and in Section 4 we present conclusions and discuss possible directions for future research.


In this paper we proposed a structured deep learning network approach for predicting the need for medically-assisted living using the survey data responses. In orderto make the algorithm scalable with respect to the type of data available we proposed a structured approach in usingseveral modules based on the type of prior information that is available. We train these modules using a deep learning approach and the national health survey and then perform optimally designed information fusion. We demonstrated that the structured MLE outperforms the other two algorithms due to the fact that for the moderately sized data it reachesstatistically optimal decision fusion. In future, an effort should be placed on evaluating the performance of structuredMLE based approach for large data sets. In addition, we planto extend deep learning network to account for multinomial nature of survey responses thus predict the level of caredneeded which can potentially be used in forecasting theoverall costs. This would enable health care providers toprovide the best possible care for given financial constraints.

About KSRA

The Kavian Scientific Research Association (KSRA) is a non-profit research organization to provide research / educational services in December 2013. The members of the community had formed a virtual group on the Viber social network. The core of the Kavian Scientific Association was formed with these members as founders. These individuals, led by Professor Siavosh Kaviani, decided to launch a scientific / research association with an emphasis on education.

KSRA research association, as a non-profit research firm, is committed to providing research services in the field of knowledge. The main beneficiaries of this association are public or private knowledge-based companies, students, researchers, researchers, professors, universities, and industrial and semi-industrial centers around the world.

Our main services Based on Education for all Spectrum people in the world. We want to make an integration between researches and educations. We believe education is the main right of Human beings. So our services should be concentrated on inclusive education.

The KSRA team partners with local under-served communities around the world to improve the access to and quality of knowledge based on education, amplify and augment learning programs where they exist, and create new opportunities for e-learning where traditional education systems are lacking or non-existent.

FULL Paper PDF file:

Predicting the Assisted Living Care Needs Using Machine Learning and health State Survey Data



A. Jeremic, D. Nikolic, M. Kostadinovic and M. S. Milicevic,




Predicting the Assisted Living Care Needs Using Machine Learning and Health State Survey Data,

Publish in

2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), Montreal, QC, Canada, 2020, pp. 5420-5423,



PDF reference and original file: Click here

+ posts

Somayeh Nosrati was born in 1982 in Tehran. She holds a Master's degree in artificial intelligence from Khatam University of Tehran.

Website | + posts

Professor Siavosh Kaviani was born in 1961 in Tehran. He had a professorship. He holds a Ph.D. in Software Engineering from the QL University of Software Development Methodology and an honorary Ph.D. from the University of Chelsea.

Website | + posts

Nasim Gazerani was born in 1983 in Arak. She holds a Master's degree in Software Engineering from UM University of Malaysia.