Data analytics problkjhvjhbjbkjbbem.pptx

kjgvljhvbljhvljhv

1. Data analytics problem Obesity Fekadesilassie Mozart

2. Introduction Obesity is a growing global health concern affecting individuals of all ages and socioeconomic backgrounds. This presentation aims to highlight the extensive obesity problem both globally and specifically in Ethiopia. By examining research results, we can gain insight into the serious implications of obesity.

3. Con’t … According to the World Health Organization (WHO), global obesity rates have tripled since 1975. In 2019, more than 1.9 billion adults were overweight, and over 650 million were obese. Obesity is associated with various health risks, including cardiovascular diseases, diabetes, and certain types of cancer.

4. Con’t … Obesity is not limited to high-income countries; it is also a growing concern in low- and middle-income countries like Ethiopia. Research conducted in Ethiopia reveals an alarming increase in obesity rates over the years.

5. Spatiotemporal distribution and determinants of overweight or obesity among urban women in Ethiopia: a multivariate decomposition analysis

6. Con’t … Study: "Prevalence of Obesity and Associated Factors among Adults in Urban Communities of Addis Ababa, Ethiopia" (EJHS) Findings: The study reported a prevalence of obesity of 7.6 % among adults in urban areas of Addis Ababa. (334,400) Overweight = 19.5% (858,000)

7. Is this a data analytics problem? YES! Principles of data analysis can be applied to this problem. To uncover factors contributing to obesity and use patterns and trends for prediction. Statistical analysis and machine learning algorithms can be applied. Predictive modeling.

8. Aim To build a predictive model using KNN Supervised machine learning algorithm. This in turn helps to predict individuals obesity class and act as a precaution mechanism. In the long run fighting obesity means decreased: Type 2 Diabetes Cardiovascular Diseases Hypertension Sleep Apnea Cancers ……. Many many more

9. The Data Data types Age: Numerical (discrete) Gender: Categorical (1 for male, 2 for female) Height: Numerical (continuous) Weight: Numerical (continuous) BMI: Numerical (continuous) PhysicalActivityLevel : Numerical (categorical scale) ObesityCategory : Categorical

10. Con’t … 2. Description The dataset consists of 1001 observations or records. It contains 7 variables or features. 48 kb .csv format 3. Missing values No missing values are apparent in the dataset.

11. Con’t … Summary description No null values All data types were: Integer or float Class label: 0 – Underweight 1 – Normal weight 2 – Over weight 3 – Obese Temporality There is no temporal data in the data set.

12. Con’t … Mean Age = 49.8 (Min = 18, Max = 79) Height = 1.70cm (Min = 1.36 cm, Max = 2.01 cm) Weight = 71.2 kg (Min = 26kg, Max = 118kg) BMI = 24.88 kg/cm^2 (Min = 8.47, Max = 50.79)

13. Con’t … Median Age = 50 Height = 169.8 cm Weight = 71.92 kg BMI = 24.69 kg/cm ^ 2 Mode Sex = M (1) Obesity class = Normal (1) My colab KNN

14. The KNN Algorithm Principles of similarity Euclidian distance The K value How to determine? Rule of thumb Overview The nearest Neighbour? Once trained – Memorize. Lazy learner

15. Some of the references 1. Spatial distribution of overweight or obesity among urban women using... | Download Scientific Diagram [Internet]. [cited 2024 May 14]. Available from: https://www.researchgate.net/figure/Spatial-distribution-of-overweight-or-obesity-among-urban-women-using-the-three_fig2_366020524 2. Spatial variation of overweight/obesity and associated factor among reproductive age group women in Ethiopia, evidence from EDHS 2016 | PLOS ONE [Internet]. [cited 2024 May 14]. Available from: https://journals.plos.org/plosone/article/figure?id=10.1371/journal.pone.0277955.g004