Subset the data to include only the employees who work in the hardware division and have a college degree. Reduce the dimensionality of the data by converting numerical variables such as age, years with the company, number of certifications, feedback scores, and salary into a smaller set of principal components that retain at least 85% of the information in the original data. Then use the principal components as predictor variables for building models for predicting the net promoter score (NPS).

 

Sales_Rep	Business	Age	Female	Years	College	Personality	Certficates	Feedback	Salary	NPS
1	Hardware	59	1	2	Yes	Diplomat	1	2.01	70200	5
2	Hardware	52	0	10	Yes	Diplomat	4	3.64	133000	10
3	Software	47	1	1	Yes	Explorer	1	3.88	52600	8
4	Hardware	61	0	2	Yes	Diplomat	3	2.70	96000	6
5	Software	39	0	1	No	Diplomat	5	3.44	122000	7
6	Hardware	28	0	6	Yes	Explorer	1	2.43	60000	6
7	Software	25	1	1	Yes	Explorer	5	3.3	68000	6
8	Hardware	51	1	10	No	Explorer	0	2.15	43800	5
9	Hardware	34	0	4	Yes	Diplomat	2	2.91	92000	7
10	Hardware	38	1	1	Yes	Explorer	5	1.23	73400	6
11	Software	53	1	11	Yes	Explorer	2	3.93	93400	8
12	Hardware	41	1	1	Yes	Sentinel	1	2.26	51800	5
13	Hardware	40	0	1	No	Diplomat	4	3.60	116000	9
14	Software	41	0	2	Yes	Explorer	3	2.17	89000	6
15	Hardware	46	1	2	Yes	Analyst	1	4.00	61800	5
16	Hardware	38	1	4	Yes	Diplomat	1	2.37	57400	4
17	Software	39	0	2	No	Sentinel	2	3.00	57000	3
18	Hardware	52	1	1	No	Sentinel	0	2.72	50600	4
19	Software	54	1	2	Yes	Explorer	5	2.04	90200	9
20	Hardware	24	0	1	Yes	Analyst	1	2.66	39000	3
21	Software	53	0	1	Yes	Explorer	2	2.42	75000	6
22	Hardware	37	1	1	Yes	Diplomat	2	3.71	96600	9
23	Software	32	1	1	Yes	Explorer	3	3.42	61600	5
24	Hardware	33	0	2	Yes	Explorer	1	1.16	60000	4
25	Software	55	1	1	Yes	Analyst	1	1.99	53000	4
26	Software	27	1	2	Yes	Diplomat	6	2.93	73600	9
27	Hardware	58	1	2	Yes	Explorer	4	3.54	73400	7
28	Software	52	1	7	Yes	Sentinel	1	3.37	46600	3
29	Software	26	0	1	Yes	Explorer	2	3.29	77000	10