Corrections_Today_January_February_2019

Office of Correctional Health

Figure 6 — BIU inmates projection – Using BIU programming data

Projected average BIU admits per month through the end of 2018

12

ACTUAL PREDICTED

10

8

Projecting average of 3 admits per month

6

4

2

0

1/1/16

2/1/16

3/1/16

4/1/16

5/1/16

6/1/16

7/1/16

8/1/16

9/1/16

1/1/17

2/1/17

3/1/17

4/1/17

5/1/17

6/1/17

7/1/17

8/1/17

9/1/17

1/1/18

2/1/18

3/1/18

4/1/18

5/1/18

6/1/18

7/1/18

8/1/18

9/1/18

10/1/16

11/1/16

12/1/16

10/1/17

11/1/17

12/1/17

10/1/18

11/1/18

12/1/18

ever-changing corrections population. Data analytics are being used in novel ways. Social network analysis of publicly available information has un- covered terror networks and organized

crime hierarchies in the community. 3 Analysis of imaging data, audio and video data coupled with location and proximity analysis will likely yield similar insights in corrections systems.

Finally, persons at high risk to com- mit or to suffer PREA incidents may be identified early, allowing proac- tive, positive action to prevent such incidents. So the question becomes, “What meaningful insights are hidden in the data you already hold?” Breakout: What are big data and data analytics? The terms “big data” and “data analytics” are becoming common- place in the fields of commerce, health care and public administra- tion. Big data often refers to the large amount of information gathered by business or governmental entities regarding individuals or groups of individuals. This information may be spread across numerous databases. Data that resides in tables in spread- sheets and databases is referred to as structured data. More often, data is

uerent predicts risk of AS (BIU Programming) and identifies residents who would benefit from BIU programs Figure 7 — Querent predicts risk of AS (BIU Programming) and identifies residents who would benefit from BIU programs

14/100

82/100

Querent

Bivariate Model

Querent > Bivariate 8.4 > 7.7

Querent 82/100

Bivariate 14/100

Querent 90/100

Bivariate 98/100

POSITIVE LIKELIHOOD RATIO

SENSITIVITY

SPECIFICITY

Proportion of never -AS residents correctly identified at classification

Proportion of AS residents correctly identified at time of classification

True Positive Rate / False Positives. This ratio helps us to be more accurate for AS residents

70 — January/February 2019 Corrections Today