MichielvEldik MSc_MADS_2021 issues

MichielvEldik / MSc_MADS_2021

0 stars 1 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

not enough data for certain combinations (especially south)

#28 MichielvEldik opened 3 years ago
0
For conditional gaussian --> refer to 'DEAL' (scutari powerpoint)

#27 MichielvEldik opened 3 years ago
0
Box_Cox transform before discretise (According to scutari); or recommended --> Hartemink because it preserves conditional stuff.

#26 MichielvEldik opened 3 years ago
0
Approximate Inference (e.g. MCMC) or exact inference?

#25 MichielvEldik opened 3 years ago
0
Descriptives and NA analyses + outlier.

#24 MichielvEldik opened 3 years ago
0
clean stuff with lapply

#23 MichielvEldik opened 3 years ago
0
Controlling for confounder that is not in the data; propensity score matching.

#22 MichielvEldik opened 3 years ago
0
Observational study, NEVER causality claims.

#21 MichielvEldik opened 3 years ago
0
Heckman correction, which variables does it apply to?

#20 MichielvEldik opened 3 years ago
0
class variable should be a FACTOR data type. Please look into it!

#19 MichielvEldik closed 3 years ago
0
Figure out a way to visualize the goegraphic part. The thesis is really not complete without it

#18 MichielvEldik opened 3 years ago
0
Create var: above median message length (1/0); yes it's dirty, but also quite nice and quick.

#17 MichielvEldik opened 3 years ago
0
Sanity check rigor

#16 MichielvEldik opened 3 years ago
0
Have another backup of data

#15 MichielvEldik closed 3 years ago
0
Bayesian Network for both message_bool AND for message_length (bef_nchar) SEPERATELY

#14 MichielvEldik opened 3 years ago
1
Customer previous activity variable

#13 MichielvEldik opened 3 years ago
0
Message bool should have a variant that is based on bef_nchars.

#12 MichielvEldik closed 3 years ago
0
in-and output comments should be corrected after structure change

#11 MichielvEldik opened 3 years ago
0
new_total_rural are NAs but can be set to 0

#10 MichielvEldik closed 3 years ago
0
nchar_perword should be "bef_nchar_perword

#9 MichielvEldik closed 3 years ago
0
Switch around P4 and P3

#8 MichielvEldik closed 3 years ago
0
Make sure that changes in P3 carry over well to P4

#7 MichielvEldik closed 3 years ago
0
Clean P4 from all the rubbish

#6 MichielvEldik closed 3 years ago
0
Discretizations of variables. (udh_indicator is maybe a good swap for urbanity)

#5 MichielvEldik closed 3 years ago
0
Product categorizations

#4 MichielvEldik closed 3 years ago
0
TF-IDF indicators (or vectorized)

#3 MichielvEldik closed 3 years ago
0
standardize quotation marks

#2 MichielvEldik closed 3 years ago
0
Fix distancjes

#1 MichielvEldik closed 3 years ago
0