Number of training examples:   10000
Number of validation examples:  5000 (for model selection)
Number of test examples:       10000

Full model selection results:
Cross validation was done on the positive acceptor model for the number of states (bw%i) 
of the full connected HMM behind the splice site.
The first part is modelled via a 'linear' HMM (just observation frequencies).
The accuracy (c:%f), error rate (w:%f), false positives (fp:%f), true positives (tp:%f) is shown for the
connected model (linear+full connected) on the validation set (only for the positive model). 
The optimal bias term (thresh*:%f) for the validation data is shown. This treshhold was used on the test set.
l with highest accuracy was chosen.

ag_pos_bw10_tr10000_r1.out:of 5000 samples (c:0.918000,w:0.082000,fp:0.060292,tp:0.883029,tresh*:-62.941453114465887)
ag_pos_bw10_tr10000_r2.out:of 5000 samples (c:0.912400,w:0.087600,fp:0.087960,tp:0.912935,tresh*:-63.717663699617738)
ag_pos_bw10_tr10000_r3.out:of 5000 samples (c:0.921800,w:0.078200,fp:0.067314,tp:0.904188,tresh*:-63.461554860448672)
ag_pos_bw10_tr10000_r4.out:of 5000 samples (c:0.917800,w:0.082200,fp:0.066153,tp:0.891304,tresh*:-63.098928106876542)
ag_pos_bw10_tr10000_r5.out:of 4999 samples (c:0.919584,w:0.080416,fp:0.072052,tp:0.905987,tresh*:-63.360715402778226)
ag_pos_bw11_tr10000_r1.out:of 5000 samples (c:0.920000,w:0.080000,fp:0.060616,tp:0.888773,tresh*:-62.999899917086772)
ag_pos_bw11_tr10000_r2.out:of 5000 samples (c:0.909400,w:0.090600,fp:0.089967,tp:0.908458,tresh*:-63.715957845850369)
ag_pos_bw11_tr10000_r3.out:of 5000 samples (c:0.919400,w:0.080600,fp:0.057929,tp:0.882723,tresh*:-63.019510476606939)
ag_pos_bw11_tr10000_r4.out:of 5000 samples (c:0.917200,w:0.082800,fp:0.075466,tp:0.905090,tresh*:-63.362025250632868)
ag_pos_bw11_tr10000_r5.out:of 4999 samples (c:0.920784,w:0.079216,fp:0.067205,tp:0.901261,tresh*:-63.275884944539811)
ag_pos_bw12_tr10000_r1.out:of 5000 samples (c:0.918600,w:0.081400,fp:0.064182,tp:0.890862,tresh*:-63.118140858000615)
ag_pos_bw12_tr10000_r2.out:of 5000 samples (c:0.911800,w:0.088200,fp:0.082274,tp:0.902985,tresh*:-63.57536153663164)
ag_pos_bw12_tr10000_r3.out:of 5000 samples (c:0.922400,w:0.077600,fp:0.068285,tp:0.907330,tresh*:-63.527145989177541)
ag_pos_bw12_tr10000_r4.out:of 5000 samples (c:0.918200,w:0.081800,fp:0.087026,tp:0.926829,tresh*:-63.797672006626939)
ag_pos_bw12_tr10000_r5.out:of 4999 samples (c:0.919584,w:0.080416,fp:0.078191,tp:0.915966,tresh*:-63.685135880374681)
ag_pos_bw13_tr10000_r1.out:of 5000 samples (c:0.918800,w:0.081200,fp:0.084279,tp:0.923760,tresh*:-63.936519265334852)
ag_pos_bw13_tr10000_r2.out:of 5000 samples (c:0.911200,w:0.088800,fp:0.086288,tp:0.907463,tresh*:-63.65641579285176)
ag_pos_bw13_tr10000_r3.out:of 5000 samples (c:0.922200,w:0.077800,fp:0.065049,tp:0.901571,tresh*:-63.366917944522797)
ag_pos_bw13_tr10000_r4.out:of 5000 samples (c:0.915200,w:0.084800,fp:0.093449,tp:0.929480,tresh*:-63.982050237746598)
ag_pos_bw13_tr10000_r5.out:of 4999 samples (c:0.921784,w:0.078216,fp:0.076252,tp:0.918592,tresh*:-63.612515130048266)
ag_pos_bw14_tr10000_r1.out:of 5000 samples (c:0.917600,w:0.082400,fp:0.077472,tp:0.909661,tresh*:-63.565145372875961)
ag_pos_bw14_tr10000_r2.out:of 5000 samples (c:0.912800,w:0.087200,fp:0.085953,tp:0.910945,tresh*:-63.724044232514515)
ag_pos_bw14_tr10000_r3.out:of 5000 samples (c:0.921400,w:0.078600,fp:0.060518,tp:0.892147,tresh*:-63.161950457490811)
ag_pos_bw14_tr10000_r4.out:of 5000 samples (c:0.915400,w:0.084600,fp:0.088953,tp:0.922587,tresh*:-63.814247136847627)
ag_pos_bw14_tr10000_r5.out:of 4999 samples (c:0.920784,w:0.079216,fp:0.072375,tp:0.909664,tresh*:-63.468215823579797)
ag_pos_bw5_tr10000_r1.out:of 5000 samples (c:0.919800,w:0.080200,fp:0.069692,tp:0.902872,tresh*:-63.470990310868231)
ag_pos_bw5_tr10000_r2.out:of 5000 samples (c:0.907000,w:0.093000,fp:0.079599,tp:0.887065,tresh*:-63.346431722027788)
ag_pos_bw5_tr10000_r3.out:of 5000 samples (c:0.916600,w:0.083400,fp:0.075081,tp:0.903141,tresh*:-63.545280507000967)
ag_pos_bw5_tr10000_r4.out:of 5000 samples (c:0.915000,w:0.085000,fp:0.088632,tp:0.920997,tresh*:-63.798303910185297)
ag_pos_bw5_tr10000_r5.out:of 4999 samples (c:0.916983,w:0.083017,fp:0.078837,tp:0.910189,tresh*:-63.674153148559952)
ag_pos_bw6_tr10000_r1.out:of 5000 samples (c:0.918800,w:0.081200,fp:0.077796,tp:0.913316,tresh*:-63.652394548103416)
ag_pos_bw6_tr10000_r2.out:of 5000 samples (c:0.908600,w:0.091400,fp:0.096321,tp:0.915920,tresh*:-63.983926860449081)
ag_pos_bw6_tr10000_r3.out:of 5000 samples (c:0.919800,w:0.080200,fp:0.073463,tp:0.908901,tresh*:-63.553467664832112)
ag_pos_bw6_tr10000_r4.out:of 5000 samples (c:0.917400,w:0.082600,fp:0.065189,tp:0.888653,tresh*:-63.04868615808968)
ag_pos_bw6_tr10000_r5.out:of 4999 samples (c:0.919184,w:0.080816,fp:0.078514,tp:0.915441,tresh*:-63.708772954927952)
ag_pos_bw7_tr10000_r1.out:of 5000 samples (c:0.920000,w:0.080000,fp:0.063857,tp:0.893995,tresh*:-63.217866590944965)
ag_pos_bw7_tr10000_r2.out:of 5000 samples (c:0.909000,w:0.091000,fp:0.085284,tp:0.900498,tresh*:-63.501955654953122)
ag_pos_bw7_tr10000_r3.out:of 5000 samples (c:0.917200,w:0.082800,fp:0.074110,tp:0.903141,tresh*:-63.563968783123883)
ag_pos_bw7_tr10000_r4.out:of 5000 samples (c:0.916200,w:0.083800,fp:0.084778,tp:0.917815,tresh*:-63.662354368819109)
ag_pos_bw7_tr10000_r5.out:of 4999 samples (c:0.921584,w:0.078416,fp:0.072375,tp:0.911765,tresh*:-63.457382014626972)
ag_pos_bw8_tr10000_r1.out:of 5000 samples (c:0.917600,w:0.082400,fp:0.074554,tp:0.904961,tresh*:-63.421047376699349)
ag_pos_bw8_tr10000_r2.out:of 5000 samples (c:0.910200,w:0.089800,fp:0.091304,tp:0.912438,tresh*:-63.924673529916873)
ag_pos_bw8_tr10000_r3.out:of 5000 samples (c:0.920600,w:0.079400,fp:0.069579,tp:0.904712,tresh*:-63.584429008752963)
ag_pos_bw8_tr10000_r4.out:of 5000 samples (c:0.915800,w:0.084200,fp:0.082852,tp:0.913574,tresh*:-63.581123608832996)
ag_pos_bw8_tr10000_r5.out:of 4999 samples (c:0.920384,w:0.079616,fp:0.081099,tp:0.922794,tresh*:-63.83253806053699)
ag_pos_bw9_tr10000_r1.out:of 5000 samples (c:0.919400,w:0.080600,fp:0.068071,tp:0.899217,tresh*:-63.263435554582131)
ag_pos_bw9_tr10000_r2.out:of 5000 samples (c:0.912200,w:0.087800,fp:0.084281,tp:0.906965,tresh*:-63.601317372595936)
ag_pos_bw9_tr10000_r3.out:of 5000 samples (c:0.922400,w:0.077600,fp:0.063107,tp:0.898953,tresh*:-63.343121733763454)
ag_pos_bw9_tr10000_r4.out:of 5000 samples (c:0.917200,w:0.082800,fp:0.092164,tp:0.932662,tresh*:-63.974918067699619)
ag_pos_bw9_tr10000_r5.out:of 4999 samples (c:0.917584,w:0.082416,fp:0.075606,tp:0.906513,tresh*:-63.605044251719846)

Best models:

ag_pos_bw11_tr10000_r1.out:of 5000 samples (c:0.920000,w:0.080000,fp:0.060616,tp:0.888773,tresh*:-62.999899917086772)
ag_pos_bw14_tr10000_r2.out:of 5000 samples (c:0.912800,w:0.087200,fp:0.085953,tp:0.910945,tresh*:-63.724044232514515)
ag_pos_bw12_tr10000_r3.out:of 5000 samples (c:0.922400,w:0.077600,fp:0.068285,tp:0.907330,tresh*:-63.527145989177541)
ag_pos_bw12_tr10000_r4.out:of 5000 samples (c:0.918200,w:0.081800,fp:0.087026,tp:0.926829,tresh*:-63.797672006626939)
ag_pos_bw13_tr10000_r5.out:of 4999 samples (c:0.921784,w:0.078216,fp:0.076252,tp:0.918592,tresh*:-63.612515130048266)

Performance of the best joined (positive with corresponding negative) models on the validation data with optimal treshhold:

1) error rate

a) for all 5 runs
    0.0274    0.0300    0.0268    0.0312    0.0270

b) mean
    0.0285

c) standard deviation
    0.0020

2) specificity ( 1 - false positives )

a) for all 5 runs
    0.9754    0.9732    0.9709    0.9785    0.9796

b) mean
    0.9755

c) standard deviation
    0.0036

3) sensitivity (true positives)

a) for all 5 runs
    0.9681    0.9652    0.9770    0.9528    0.9622

b) mean
    0.9651

c) standard deviation
    0.0088