Alexander-Schiendorfer

predictions = model.predict(test_images)
predictions = np.argmax(predictions, axis=1)
conf_matrix = confusion_matrix(np.argmax(test_labels, axis=1), predictions)

# just for better presentation - but store original values
default_figsize = plt.rcParams['figure.figsize'] 
default_dpi = plt.rcParams['figure.dpi']

plt.rcParams['figure.figsize'] = [19, 12]
plt.rcParams['figure.dpi'] = 150

Alexander-Schiendorfer

 def compare_nearest_neighbors(self):

        """ get nearest neighbor for every wrong labeled datapoint and compare feature values in one DataFrame """

        results = []
        for index, row in self.wrong_predictions.iterrows():
            query_data_point = row[:-2]
            query_data = row[:]
            data_points = self.X_predictions.iloc[:, :-2]

Alexander-Schiendorfer

# training für 10 epochen
history = model.fit(train_images, train_labels, batch_size=batch_size, 
                    shuffle=True, epochs=n_epochs, validation_data = (test_images, test_labels))

import pandas as pd
pd.DataFrame(history.history).plot(figsize=(8,5))
plt.grid(True)
plt.gca().set_ylim(0, 1)
plt.show()

Alexander-Schiendorfer

# first make sure to save the history ... 

# let's actually fit our model 
history = model.fit(train_images, train_labels, epochs=num_epochs, batch_size=batch_size)

# then, visualize the history
import pandas as pd

pd.DataFrame(history.history).plot(figsize=(8, 5))
plt.grid(True)

Alexander-Schiendorfer

nCarTypes = 6;
nOptions = 5;

requires = array2d(CarTypes, Options, [
  true, false, true, true, false,
  false, false, false, true, false,
  false, true, false, false, true,
  false, true, false, true, false,
  true, false, false, false, false,
  true, true, false, false, false]);

Alexander-Schiendorfer

```matlab
include "classic_o.mzn"; % output of minibrass
include "soft_constraints/pvs_gen_search.mzn"; % for generic branch and bound

% the basic, "classic" CSP 
set of int: NURSES = 1..3;
int: day = 1; int: night = 2; int:off = 3;
set of int: SHIFTS = {day,night,off};

array[NURSES] of var SHIFTS: n;

Alexander-Schiendorfer

% A really simplistic over-constrained model
% X: {x,y,z} D_i = {1,2,3}, i in X
%    * c1: x + 1 = y 
%    * c2: z = y + 2
%    * c3: x + y <= 3
% -------------------------------------------

include "soft-constraints/soft_constraints.mzn"; % model additions for soft constraint business
include "soft-constraints/spd_worse.mzn"; % the actual isBetter predicate
include "soft-constraints/tpd_worse.mzn"; % the actual isBetter predicate

Alexander-Schiendorfer

% a very minimal example to provoke isa: nullptr bug
% ---------------------------------
include "minisearch.mzn"; % include the search minisearch lite

set of int: range = 1..5;
var set of range: x;

predicate onlyEvens(var set of range: y) = 
(
 forall(r in y) (r mod 2 = 0)

Alexander-Schiendorfer

% a very minimal example to provoke isa: nullptr bug
% ---------------------------------

set of int: range = 1..5;
var set of range: x;

predicate onlyEvens(var set of range: y) = 
(
 forall(r in y) (r mod 2 = 0)
);

Alexander-Schiendorfer

% a very minimal example to provoke isa: nullptr bug
% ---------------------------------

set of int: range = 1..5;
var set of range: x;

predicate onlyEvens(var set of range: y) = 
(
 forall(r in y) (r mod 2 = 0)
);