] # if SPDhits: variables += ['SPDhits'] # if isolationc: variables += ['isolationc'] # if IPSig: variables += ['IPSig'] # if IP: variables += ['IP'] if np.round(hyperparameters.pop("SPDhits")): variables += ["SPDhits"] if np.round(hyperparameters.pop("isolationc")): variables += ["isolationc"] if np.round(hyperparameters.pop("IPSig")): variables += ["IPSig"] if np.round(hyperparameters.pop("IP")): variables += ["IP"] hyperparameters["numFeatures"] = len(variables) data = loadTrainingData(variables, rng, valPart=0.3, verbose=True) optAlgs = ["RMSProp", "adaDelta", "vSGDfd"] patience = 12000 validationFrequency = 1 batchSize = int(data[0].get_value(borrow=True).shape[0] / 12.0) # totNumSamples = 67553 numTrainBatches = data[0].get_value(borrow=True).shape[0] / batchSize numEpochs = int(patience / numTrainBatches) * 15 hyperparameters["optAlg"] = optAlgs[1] errorDict = {} AUCDict = {} mlp = initalizeModel(**hyperparameters) score = trainWithEarlyStopping( mlp,
import theano folder = "saves\\16L500LS0.85FD0.9D0SPD1ISO1IPS1IP5.00C0.00L10.50L20.75RHO1e-10EPS13002242\\" hyperparameters = {'c': 5, 'SPDhits': 0.49, 'dropout': 0.9, 'firstLayerDropout': 0.85, 'layerSize': 500, 'rho': 0.75, 'epsilon': 1e-10, 'numLayers': 16, 'lambda1': 1e-4, 'lambda2': 0.5} hyperparameters["isolationc"] = 1 hyperparameters["IPSig"] = 1 hyperparameters["IP"] = 1 rng = np.random.RandomState(123) hyperparameters["rng"] = rng variables = ['FlightDistance','FlightDistanceError', 'LifeTime', 'VertexChi2','pt','dira','DOCAone', 'DOCAtwo','DOCAthree','IP_p0p2','IP_p1p2', 'isolationa', 'isolationb', 'isolationd', 'isolatione', 'isolationf', 'iso', 'CDF1', 'CDF2', 'CDF3', 'ISO_SumBDT', 'p0_IsoBDT', 'p1_IsoBDT', 'p2_IsoBDT', 'p0_track_Chi2Dof','p1_track_Chi2Dof', 'p2_track_Chi2Dof','p0_pt','p0_p','p0_eta', 'p0_IP','p0_IPSig','p1_pt','p1_p', 'p1_eta','p1_IP','p1_IPSig','p2_pt','p2_p','p2_eta','p2_IP', 'p2_IPSig'] if np.round(hyperparameters.pop('SPDhits')): variables += ['SPDhits'] if np.round(hyperparameters.pop('isolationc')): variables += ['isolationc'] if np.round(hyperparameters.pop('IPSig')): variables += ['IPSig'] if np.round(hyperparameters.pop('IP')): variables += ['IP'] hyperparameters["numFeatures"] = len(variables) optAlgs = ["RMSProp", "adaDelta", "vSGDfd"] hyperparameters["optAlg"] = optAlgs[1] Xtrain, Ytrain, Xval, Yval, mu, sig, AUCindices = loadTrainingData(variables, rng, valPart = 0.3, verbose = True) model = initalizeModel(**hyperparameters) model.loadModel("", folder, verbose = True) generateSubmission(model, variables, mu, sig)