Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chapters 13 - 15 (Build Candidate Models (Advanced options (random,…
Chapters 13 - 15
Build Candidate Models
starting the process
select the target feature.
LogLoss
Advanced options
random
stratified
Partition Feature
Train, validation, holdout sample
Group approach
Tournament Round
Feature understanding and selection
Data types
indicates the nature of the data inside of a feature.
descriptive statistics
mean
min
max
median
std dev
feature content
missing data
Startup Process
Uploading data
Stick to smaller sample data sets while learning
Data Robot will accept comma separated files
Will read .CSV files
May be better to store data as .tsv
Excel files are best bet when possible
URL link to dataset
compressing data may speed things up
DATA Robot limits
Dataset must be 100 rows or more
Less than or equal to 20000 columns
less than or equal to 1.5 gigabytes