Data
sf-police-incidents

sf-police-incidents

active ARFF ODC Public Domain Dedication and Licence (PDDL) Visibility: public Uploaded 03-04-2020 by Florian Pargent
0 likes downloaded by 0 people , 0 total downloads 0 issues 0 downvotes
Issue #Downvotes for this reason By


Loading wiki
Help us complete this description Edit
Incident reports from the San Franciso Police Department between January 2003 and May 2018, provided by the City and County of San Francisco. The dataset was downloaded on 05.11.2018. from [https://data.sfgov.org/Public-Safety/Police-Department-Incident-Reports-Historical-2003/tmnf-yvry]. For a description of all variables, checkout the homepage of the data provider. The original data was published under ODC Public Domain Dedication and Licence (PDDL) [https://opendatacommons.org/licenses/pddl/1.0/]. As target, the binary variable 'ViolentCrime' was created. A 'ViolentCrime' was defined as 'Category' %in% c('ASSAULT', 'ROBBERY', 'SEX OFFENSES, FORCIBLE', 'KIDNAPPING') | 'Descript' %in% c('GRAND THEFT PURSESNATCH', 'ATTEMPTED GRAND THEFT PURSESNATCH'). Additional date and time features 'Hour', 'DayOfWeek', 'Month', and 'Year' were created. The original variables 'Category', 'Descript', 'Date', 'Time', 'Resolution', 'Location', and 'PdId' were removed from the dataset. One record which contained the only missing value in the variable 'PdDistrict' was removed from the dataset. Using this dataset for machine learning was inspired by Nina Zumel's blogpost [http://www.win-vector.com/blog/2012/07/modeling-trick-impact-coding-of-categorical-variables-with-many-levels/]. Note that incidents consist of multiple rows in the dataset when the crime belongs to more than one 'Category', which is indicated by the ID variable 'IncidntNum' (ignored by default). For this version, the majority class was downsampled to achieve a balanced classification task. Unused factor levels were dropped. The numeric features 'X' and 'Y' were removed to increase the importance of the high cardinal factorial features

7 features

ViolentCrime (target)nominal2 unique values
0 missing
Hournumeric24 unique values
0 missing
DayOfWeeknominal7 unique values
0 missing
Monthnominal12 unique values
0 missing
Yearnominal16 unique values
0 missing
PdDistrictnominal10 unique values
0 missing
Addressnominal21838 unique values
0 missing

19 properties

538638
Number of instances (rows) of the dataset.
7
Number of attributes (columns) of the dataset.
2
Number of distinct values of the target attribute (if it is nominal).
0
Number of missing values in the dataset.
0
Number of instances with at least one value missing.
1
Number of numeric attributes.
6
Number of nominal attributes.
269319
Number of instances belonging to the most frequent class.
50
Percentage of instances belonging to the least frequent class.
269319
Number of instances belonging to the least frequent class.
1
Number of binary attributes.
14.29
Percentage of binary attributes.
0
Percentage of instances having missing values.
1
Average class difference between consecutive instances.
0
Percentage of missing values.
0
Number of attributes divided by the number of instances.
14.29
Percentage of numeric attributes.
50
Percentage of instances belonging to the most frequent class.
85.71
Percentage of nominal attributes.

1 tasks

0 runs - estimation_procedure: 50 times Clustering
Define a new task