Linear-Regression- Predicting Life Expectancies of Different Countries

This is a simple linear regression and Normal equation implementation to calculate the average life expectancies of different countries. The life expectancies are caluclated by using the GDP per capita, Exponential Pollution scale index, and population density per km^2 as the parameters. The above constraints are selected to demonstrate their impact on the health of the citizens of a respected country. The purpose of this project however is not to create an accurate representation, as there are many other design constraints that need to be taken into account like the efficiency of the ruling government, climatic conditions, healthcare development, lifestyle factors like exercise rate, food consumption etc.

Motivation: This project is solely intended for Algorithm learning purposes.

Tools Used:

I have used Octave Version 5.2 to do this project, but the same can also be done using MATLAB, with some possibility of minor changes in the syntax, please check if you face any difficulties while using the code in MATLAB.

Files:

1. D1.txt - Prepared Data file - Text file.
2. Life_Expec_Calc.m - main file that runs the program.
3. costFunction.m - Code for the cost function calculator function.
4. featureNorm.m - Code for the feature normalization function.
5. gD_Calc.m - Function to calculate the Gradient descent.
6. normalEquation.m - Function to calculate the prediction using the normal equation.

Steps:

1)Data Preparation:

The data for this project was collected from different sources as a single source with all neccessary data was unavailable. The data sources are:

1. https://www.who.int
2. https://www.worldometers.info/
3. https://www.numbeo.com
4. https://en.wikipedia.org
5. https://data.worldbank.org

The data that was prepared is as follows:

Sl No.	Country	Density (P/Km²)	GDP Per capita ($)	Exp Pollution Scale Index	Life Expectancy
1	Afghanistan	60	538.00	160	63.2
2	Albania	105	4,521.00	138	78
3	Algeria	18	4,048.00	114	77.1
4	Andorra	164	39,128.00	55	83.3
5	Angola	26	4,096.00	176	63.1
6	Antigua and Barbuda	223	15,825.00	116	76.5
7	Argentina	17	14,508.00	87	76.6
8	Armenia	104	3,918.00	108	76
9	Australia	3	53,831.00	37	83
10	Austria	109	47,261.00	33	81.6
11	Azerbaijan	123	4,139.00	130	71.4
12	Bahrain	2,239	23,715.00	129	75.8
13	Bangladesh	1,265	1,564.00	152	74.3
14	Barbados	668	16,328.00	49	76
15	Belarus	47	5,762.00	74	74.8
16	Belgium	383	43,325.00	84	81.4
17	Belize	17	4,957.00	96	74.4
18	Bhutan	20	3,391.00	93	73.1
19	Bolivia	11	3,351.00	128	72.1
20	Bosnia and Herzegovina	64	5,387.00	108	76.8
21	Brazil	25	9,881.00	93	75.9
22	Bulgaria	64	8,197.00	116	75.1
23	Cambodia	95	1,384.00	138	70.1
24	Cameroon	56	1,422.00	162	62.4
25	Canada	4	44,841.00	46	82.2
26	Chile	26	15,001.00	141	80.7
27	China	153	8,612.00	148	77.4
28	Colombia	46	6,429.00	112	79.3
29	Congo (Congo-Brazzaville)	16	1,703.00	175	64.7
30	Costa Rica	100	11,573.00	91	80.8
31	Croatia	73	13,200.00	50	78.6
32	Cuba	106	8,541.00	114	77.8
33	Cyprus	131	18,695.00	93	83.1
34	Czechia (Czech Republic)	139	20,291.00	60	79.1
35	Denmark	137	57,545.00	32	81.3
36	Dominican Republic	225	7,223.00	132	72.8
37	Ecuador	71	6,214.00	99	78.4
38	Egypt	103	2,441.00	149	71.8
39	Estonia	31	20,170.00	32	78.9
40	Ethiopia	115	757.00	136	68.7
41	Fiji	49	5,768.00	90	68
42	Finland	18	45,778.00	20	81.6
43	France	119	39,827.00	70	82.5
44	Georgia	57	3,762.00	125	73.3
45	Germany	240	44,680.00	45	81.7
46	Ghana	137	2,026.00	159	66.3
47	Greece	81	19,214.00	90	81.1
48	Guatemala	167	4,471.00	127	72
49	Haiti	414	766.00	155	64.1
50	Hong Kong	6,670	46,733.00	120	71.9
51	Honduras	89	2,405.00	140	85
52	Hungary	107	14,364.00	83	76.4
53	Iceland	3	73,233.00	27	82.3
54	India	464	1,980.00	141	70.8
55	Indonesia	151	3,837.00	115	71.3
56	Iran	52	5,628.00	138	77.3
57	Iraq	93	5,114.00	128	72.4
58	Ireland	72	69,727.00	56	81.8
59	Israel	400	42,852.00	103	82.6

The data was then extracted and made into comma seperated format in the form of a text file 'D1.txt"