ISY503 Intelligent Systems Report Sample

ISY503 Intelligent Systems Report

Task Instructions

To complete this assessment task you must write a case study report with the following sections:

• Introduction: This section should introduce the case study you have been given, and highlight the significance of the problem it seeks to address.

• Background: This section should provide sufficient background information on, and explain the application of, the intelligent system including machine learning models and methods.

• Method: Elaborate here on the method(s) used and explain how the research was undertaken. You should include the source of data used in the case study, and identify any ethical issues that may have arisen upon its use (e.g., medical history of patients).

• Results: What was the outcome of the case study?

• Discussion: In this section you should explain the relevant and significance of your chosen study, and you should identify obstacles restricting the intelligent system. You should also mention any constraints reported in the article.

• Recommendations: This section should include critical perspectives and recommendations to improve or enhance the system.

• References: You should support your report with additional peer-reviewed journal articles.

These should be in appropriate APA style.

Solution

Introduction

Machine learning is one of the core parts of Artificial Intelligence that enables machines for recognising the patterns from the past experiences and existing data. In order to identify behavioural phenotypes in people with autism spectrum disorder (ASD) computer visions as well as machine learning are applied. This is the main focus of the case study “Computer Vision and Behavioral Phenotyping: An Autism Case Study” by Guillermo Sapiro, Jordan Hashemi and Geraldine Dawson (Sapiro, Hashemi & Dawson, 2019). In order to potentially improve outcomes through early intervention this research is important because it addresses the growing need for early accurate and non-invasive diagnostic tools for ASD. Hence a major unmet need in ASD in exacting and in developmental as well as neurodegenerative disorders in wide-ranging aspect is the development of computational approaches to standardized purposive behavioural evaluation. Computer vision tools include an affect coding pattern of movement and gaze as well as attention monitoring.

Background

The complex neurodevelopment disorder for MBA assignment expert known as autism spectrum disorder is typified by difficulties with communication as well as social interaction and also repetitive behaviours. The long behavioural evaluations used in traditional ASD diagnostic procedures are carried out by qualified experts. This can also be subjective and also very much time-consuming. A possible substitute is provided by the progress of intelligent systems specifically in the areas of machine learning and computer vision (Raj & Masood, 2020). These devices specifically have the ability to analyze visual data in order to categorize behavioural cues that may be delicate as well as suggestive and subtle of ASD. The authors of this case study mainly employ techniques of computer vision to scrutinize children’s behaviour captured on video recording. Certain behavioural patterns such as facial expressions as well as gestures and also abnormal gaze patterns are trained into machine learning models in order to recognize them as signs of ASD. This particular strategy seeks to offer a more objective and scalable method for screening as well as diagnosing for ASD. Additionally, from the point of view of Sapiro, Hashemi & Dawson (2019), it is stated that the asymmetric position of the arms of toddlers is also considered here. This is also effective in identifying the ASD.

Method

In this study, it has been found that there are numerous factors to diagnose as well as screen for ASD. For this reason, this particular study has focused on the secondary analysis. Different previous studies are used to establish the argument in this specific research. Behavioural phenotype is taken into consideration in this regard (Thabtah et al. 2022). The efficiency of the computer vision is analysed in this study to evaluate the factors of ASD. It can be stated that the needs for the computer vision for the coding of the behaviour of children are understood to identify the ASD in an effective manner. Digital behaviour measurement is also analysed in this study collectively to treat children who are having neurodevelopment disorders.

The study follows a qualitative method to discuss the aspects of ASD in children on an in-depth note. A detailed analysis of the attention as well as gaze of children for recognising the ASD is discussed here. Different articles are taken for analysing the factors of ASD considering the computer vision as well as machine learning. Additionally, a quantitative method like motor analysis is also described in this particular study. This is an objective as well as a numeric method to identify the assessment of uncharacteristic motor or movement in ASD (Sapiro, Hashemi & Dawson, 2019). This method is moderately effective for the researcher to understand the aspects as well as different factors associated with the ASD in children. On the contrary, an affect analysis is also mentioned in this research study. Dissimilarities in affect as well as emotion are an imperative biomarkers for frequent neurodegenerative as well as developmental disorders and have been studied widely in ASD also. The main data collection method used in this study is video recording to recognise the gaze as well as behavioural aspects of children. Additionally, patterns of movement as well as farcical expression are also recorded to study the factors of ASD. The usage of a machine learning model for categorising the behaviour of children as the indicator of the ASD is critically discussed in this study (Sapiro, Hashemi & Dawson, 2019).

The ethical aspects are followed in this study by considering the consent as well as privacy protocol for recording video of children. The informed consents are obtained from the guardians of the children. The names of children are anonymized here. The recorded videos are annotated with the makers of behaviour by the experts.

Result

The study proved that the computer vision-based system could recognize ASD-related behavioural patterns with accuracy. For this reason, it is mandatory to use the computer vision here. It helps to measure the behaviour of children accurately (Bala et al. 2022) It can be feasible to focus on the information system utilisation to understand the behavioural patterns of children. Based on behavioural approaches, it is required to assess the ASD considering the motor analysis as well as affect analysis (Sapiro, Hashemi & Dawson, 2019). Phenotypes the machine learning models have been able to detect children who are living with ASD with high precision from characteristically developing children. Based on the findings of this particular study it appears that the intelligent system could be a useful tool for early ASD screening as well as diagnosing by identifying the behavioural pattern, gaze pattern as well as facial expression of children (Sapiro, Hashemi & Dawson, 2019). The video recording helps to evaluate the behavioural patterns of children. The researchers have focused on the gaze pattern, behaviour as well as movements of children to understand the ASD.

Discussion

In this context, a scalable screening tool is required to diagnose ASD among children. The screening tool is objective as well as non invasive. This method of diagnosis is quite helpful for the experts to understand the factors associated with the ASD. However, in the case of the traditional method of diagnosis, there is an unavailability of the trained professionals for diagnosing ASD (Sapiro, Hashemi & Dawson, 2019). For this specific reason, computer vision is significant as well as systematic. In the case of the traditional method of diagnosing ASD, the experts use the subjective method in this context. On the other hand, the computer vision helps to make the entire method objective in nature. Affect analysis as well as motor analysis are crucial for screening as well as diagnosing ASD among children in an objective way. The mentioned system of intelligence helps to overcome the challenges as well as obstacles of subjective assessment of behaviour (Sapiro, Hashemi & Dawson, 2019).

However, in the case of the intelligent system, it has been recognised that there are some issues that occur while identifying the behavioural assessment. In the presentation of ASD, there could be variability while using the intelligent system (Zhu et al. 2023). It may affect the model of generalizability. Additionally, ethical consideration ought to be concentrated in this regard. There could be a chance of data breach occurrence or privacy concerns incidence (Sapiro, Hashemi & Dawson, 2019). The entire discussion focuses on the effectiveness of the intelligent system as well as the challenges in implementing and using the intelligent system for the purpose of screening as well as diagnosing of ASD.

Recommendation

Some suggestions can be given in this regard. These are outlined below:

Expansion of the dataset

It is suggested to expand the entire data set for managing the information systematically. It helps to deal with the obstacle of the implementation and utilisation of the intelligent system. This is useful for managing the assessment of the behaviour of children with ASD (Cavus et al. 2021).

Improvement of the extraction of feature

The features of the intelligent system should be extracted on a proper note. It is needed to improve the feature extraction to meet the standard of the diagnosis as well as screening of ASD. The machine learning for the prediction of the behaviour of the children who are living with ASD is required. It is possible through the development of the extraction of features of the system (Vakadkar, Purkayastha & Krishnan, 2021).

Robust validation

From the study of Thabtah & Peebles (2020), it can be understood that the reliability of the intelligent system ought to be taken into account here. The system needs to be reliable as well as robust in nature to manage the data properly.

Reference

Bala, M., Ali, M. H., Satu, M. S., Hasan, K. F., & Moni, M. A. (2022). Efficient machine learning models for early stage detection of autism spectrum disorder. Algorithms, 15(5), 166. DOI: https://doi.org/10.3390/a15050166

Cavus, N., Lawan, A. A., Ibrahim, Z., Dahiru, A., Tahir, S., Abdulrazak, U. I., & Hussaini, A. (2021). A systematic literature review on the application of machine-learning models in behavioral assessment of autism spectrum disorder. Journal of Personalized Medicine, 11(4), 299. DOI: https://doi.org/10.3390/jpm11040299

Raj, S., & Masood, S. (2020). Analysis and detection of autism spectrum disorder using machine learning techniques. Procedia Computer Science, 167, 994-1004. DOI: https://doi.org/10.1016/j.procs.2020.03.399

Sapiro, G., Hashemi, J., & Dawson, G. (2019). Computer vision and behavioral phenotyping: an autism case study. Current Opinion in Biomedical Engineering, 9, 14-20.DOI: https://doi.org/10.1016/j.cobme.2018.12.002

Thabtah, F., & Peebles, D. (2020). A new machine learning model based on induction of rules for autism detection. Health informatics journal, 26(1), 264-286. DOI: https://doi.org/10.1177/1460458218824711

Thabtah, F., Spencer, R., Abdelhamid, N., Kamalov, F., Wentzel, C., Ye, Y., & Dayara, T. (2022). Autism screening: An unsupervised machine learning approach. Health Information Science and Systems, 10(1), 26. DOI: 10.1007/s13755-022-00191-x

Vakadkar, K., Purkayastha, D., & Krishnan, D. (2021). Detection of autism spectrum disorder in children using machine learning techniques. SN computer science, 2, 1-9. DOI: https://doi.org/10.1007/s42979-021-00776-5

Zhu, F. L., Wang, S. H., Liu, W. B., Zhu, H. L., Li, M., & Zou, X. B. (2023). A multimodal machine learning system in early screening for toddlers with autism spectrum disorders based on the response to name. Frontiers in Psychiatry, 14, 1039293. DOI: https://doi.org/10.3389/fpsyt.2023.1039293