I have a project using R programming or Python with report
a) the description of attributes (columns) e.g. the number of discrete and continuous attributes
(make factors from discrete ones, get rid of useless ones or remove one attribute from each pair
of strongly correlated attributes). Find or make your target class attribute.
b) the description of used classification methods and their validation process (changing their
parameters e.g. minsplit, cp, k-fold crossvalidation, boosting, bagging, adding cost matrices to
them, modifing input data e.g. standardization i.e. standard score, normalization, removing NA
and so on).
c) the validation process summary including ROC plots and their comparison.
d) You may use some clustering methods for not labeled datasets (without an obvious target
attribute). After this operation you can learn classifiers utilising input data and obtained cluster
e) Enclose please your R code with comments and links to data used in R code (if it was
changed before [login to view URL] then dropbox links to your new data).
38 freelancere byder i gennemsnit $167 på dette job
I am an expert in data science, I've done a lot of projects in the field involving data pre-processing, data visualization, building predictive and clustering models, contact me for further discussion.