enhance DAE/SDAE

[jeongyoonlee]

• add transfer learning with the pretrained_model input argument
• allow to set the learning_rate in __init__()
• add a test for transfer learning between DAE/SDAE

[codecov]

Codecov Report

Merging #63 (84a9fce) into master (018ac04) will increase coverage by 5.55%.
The diff coverage is 79.15%.

@@            Coverage Diff             @@
##           master      #63      +/-   ##
==========================================
+ Coverage   51.04%   56.60%   +5.55%     
==========================================
  Files          19       20       +1     
  Lines        1050     1302     +252     
==========================================
+ Hits          536      737     +201     
- Misses        514      565      +51     

Impacted Files	Coverage Δ
kaggler/data_io.py	0.00% <0.00%> (ø)
kaggler/ensemble/linear.py	0.00% <ø> (ø)
kaggler/feature_selection/feature_selection.py	0.00% <0.00%> (ø)
kaggler/preprocessing/numerical.py	30.76% <33.33%> (-5.24%)	⬇️
kaggler/preprocessing/autoencoder.py	77.50% <77.50%> (ø)
kaggler/preprocessing/categorical.py	93.62% <97.95%> (+0.53%)	⬆️
kaggler/__init__.py	100.00% <100.00%> (ø)
kaggler/metrics/classification.py	29.16% <100.00%> (-7.88%)	⬇️
kaggler/metrics/regression.py	37.50% <100.00%> (-6.95%)	⬇️
kaggler/model/automl.py	77.17% <100.00%> (+0.25%)	⬆️
... and 4 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5274ac3...84a9fce. Read the comment docs.