Commits · cea93717c4b511cfbafb2b432925b73a3be5ad42 · Jon Lachmann / sgd · GitLab

Aug 12, 2015
- remove #include data_point header in learning rates · cea93717
  Dustin Tran authored Aug 12, 2015
  
  cea93717
Aug 11, 2015
- move delta private var to implicit_sgd derived class only · 0fcf8896
  Dustin Tran authored Aug 11, 2015
  
  0fcf8896
- remove need for grad_func(); let gradient potentially use all data · a3ac85db
  Dustin Tran authored Aug 11, 2015
  
  a3ac85db
Aug 10, 2015
- remove verbose comments · aaadd7b2
  Dustin Tran authored Aug 10, 2015
  
  aaadd7b2
Jul 31, 2015
- remove indexing into convergence list element · 6381f48c
  Dustin Tran authored Jul 11, 2015
  
  6381f48c
Jul 06, 2015
- always output converged attribute · dc7546af
  Dustin Tran authored Jul 02, 2015
  
  dc7546af
Jul 01, 2015
- implement sgd.function; improve documentation · 3dae987f
  Dustin Tran authored Jul 01, 2015
  
  3dae987f
Jun 30, 2015
- update README.md · eac49a5d
  Dustin Tran authored Jun 30, 2015
  
  eac49a5d
- update description of package · c44ead7f
  Dustin Tran authored Jun 30, 2015
  
  c44ead7f
- minor change in documentaiton of start argument; fix #58 · edce7abb
  Dustin Tran authored Jun 30, 2015
  
  edce7abb
- update README.md · 84b357be
  Dustin Tran authored Jun 30, 2015
  
  84b357be
- add classical momentum as new sgd method; #59 · a65fc476
  Dustin Tran authored Jun 30, 2015
  
  a65fc476
- add nesterov momentum as new sgd method; #59 · dacbb3c8
  Dustin Tran authored Jun 30, 2015
  
  dacbb3c8
- add demo for estimating equations · 70e14678
  Dustin Tran authored Jun 30, 2015
  
  70e14678
- fix class printing for ee model · 6ad8b1dd
  Dustin Tran authored Jun 30, 2015
  
  6ad8b1dd
- update README.md · f7262569
  Dustin Tran authored Jun 30, 2015
  
  f7262569
- split number of params from number of features; change ee_model constructor · 66c9717a
  Dustin Tran authored Jun 30, 2015
  
  66c9717a
Jun 29, 2015
- remove need for model.attrs in model.control · b482a131
  Dustin Tran authored Jun 29, 2015
  
  b482a131
- set ddim learning rate to store only a vector, not a matrix · 5911b13f
  Dustin Tran authored Jun 29, 2015
  
  5911b13f
- complete documentation for sgd output · 5ce3a084
  Dustin Tran authored Jun 29, 2015
  
  5ce3a084
- allow size to be argument in sgd_control · 0a260b70
  Dustin Tran authored Jun 29, 2015
  
  0a260b70
- fix poisson regression demo since offset no longer implemented · a15eea29
  Dustin Tran authored Jun 29, 2015
  
  a15eea29
- add explicit_sgd and implicit_sgd as subclasses of a base_sgd · bf1c96d6
  Dustin Tran authored Jun 29, 2015
  
  bf1c96d6
- split experiment argument into model_control and sgd_control · 6bc3c629
  Dustin Tran authored Jun 29, 2015
  
  6bc3c629
- refactor learning rate value to have private members with getter methods · 93d7190d
  Dustin Tran authored Jun 29, 2015
  
  93d7190d
- add error to verbose messages · 49988c1e
  Dustin Tran authored Jun 29, 2015
  
  49988c1e
- split experiment class into sgd and model-affiliated classes · 835cb471
  Dustin Tran authored Jun 29, 2015
  
  835cb471
- set lambda1, lambda2 as model.control parameters · bb230d6d
  Dustin Tran authored Jun 29, 2015
  
  bb230d6d
Jun 28, 2015
- rename experiment to model; first steps on partitioning model and sgd attributes · 6f326d2c
  Dustin Tran authored Jun 28, 2015
  
  6f326d2c
- move experiment conversions to base_experiment; add documentation · 22caf829
  Dustin Tran authored Jun 28, 2015
  
  22caf829
- remove d from learning rate operator() argument; add documentation · ba01a78e
  Dustin Tran authored Jun 28, 2015
  
  ba01a78e
- add documentation to sgd algorithms; add const references to data_set · f5160607
  Dustin Tran authored Jun 28, 2015
  
  f5160607
- add documentation to data classes; move boost::timer outside of data_set · 18dbdc1a
  Dustin Tran authored Jun 28, 2015
  
  18dbdc1a
- remove init() and covariance() methods from data_set; clean main code in sgd.cpp · 8d8d10bf
  Dustin Tran authored Jun 28, 2015
  
  8d8d10bf
Jun 27, 2015
- remove DEBUG and __R__ flags · 03fcf8d4
  Dustin Tran authored Jun 27, 2015
  
  03fcf8d4
- remove redundant basedef typedefs · 45c42c6b
  Dustin Tran authored Jun 27, 2015
  
  45c42c6b
- remove unimplemented usage of 'offset' term · 0f5791b5
  Dustin Tran authored Jun 26, 2015
  
  0f5791b5
- reduce need of init learning rate methods, shared_ptr types · 332c545f
  Dustin Tran authored Jun 26, 2015
  
  332c545f
- convert learning rates to operator() method · 7acf081d
  Dustin Tran authored Jun 26, 2015
  
  7acf081d
Jun 26, 2015
- temp · 3ee50330
  Dustin Tran authored Jun 26, 2015
  
  3ee50330