svd vs timesvd(with only 1 time bin)

[Ankush Sachdeva]

If I run a dataset with all timebins 1(i.e all ratings at in one time bin), shouldnt the svd and timesvd give the same RMSE.
I tried doing this but timesvd is giving RMSE of 0.96 and svd is giving 0.86 on same parameters.

I am trying on the movielens dataset with 1/10 cross validation set.(files attached)

The parameters are

./toolkits/collaborative_filtering/timesvdpp --training=out.txt --kfold_cross_validation=10 --kfold_cross_validation_index=3 --minval=1 --maxval=5 --max_iter=600 --quiet=1 --D=65  --lrate=0.01 --beta=1 --gamma=1

and

./toolkits/collaborative_filtering/svdpp --training=out_notime.txt --kfold_cross_validation=10 --kfold_cross_validation_index=3 --biassgd_lambda=0.01 --biassgd_gamma=0.01 --minval=1 --maxval=5 --max_iter=600 --quiet=1 --D=65 --svdpp_item_bias_step=0.01 --svdpp_user_bias_step=0.01 --svdpp_user_factor_step=0.01 --svdpp_user_factor2_step=0.01 --svdpp_item_bias_reg=1 --svdpp_user_bias_reg=1 --svdpp_user_factor_reg=1 --svdpp_user_factor2_reg=1 --svdpp_step_dec=0.90