Example of Health State Classification¶

Cloning the damavand repository¶

In [ ]:

Copied!

!git clone https://github.com/pydamavand/damavand
!git clone https://github.com/pydamavand/damavand

fatal: destination path 'damavand' already exists and is not an empty directory.

In [2]:

Copied!

!pip install -r damavand/requirements.txt
!pip install -r damavand/requirements.txt

Requirement already satisfied: certifi==2024.7.4 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 1)) (2024.7.4)
Requirement already satisfied: charset-normalizer==3.3.2 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 2)) (3.3.2)
Requirement already satisfied: idna==3.7 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 3)) (3.7)
Requirement already satisfied: numpy==1.26.4 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 4)) (1.26.4)
Requirement already satisfied: pandas==2.1.4 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 5)) (2.1.4)
Requirement already satisfied: python-dateutil==2.9.0.post0 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 6)) (2.9.0.post0)
Requirement already satisfied: pytz==2024.1 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 7)) (2024.1)
Requirement already satisfied: rarfile==4.2 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 8)) (4.2)
Requirement already satisfied: requests==2.32.3 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 9)) (2.32.3)
Requirement already satisfied: scipy==1.13.1 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 10)) (1.13.1)
Requirement already satisfied: six==1.16.0 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 11)) (1.16.0)
Requirement already satisfied: tzdata==2024.1 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 12)) (2024.1)
Requirement already satisfied: urllib3==2.2.2 in /usr/local/lib/python3.10/dist-packages (from -r damavand/requirements.txt (line 13)) (2.2.2)

Importings¶

In [23]:

Copied!





from damavand.damavand.datasets.downloaders import read_addresses, ZipDatasetDownloader
from damavand.damavand.datasets.digestors import SEU
from damavand.damavand.signal_processing import *
from damavand.damavand.utils import *

import pandas as pd
from scipy.stats import skew, kurtosis

from sklearn.preprocessing import LabelEncoder, LabelBinarizer, MinMaxScaler
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier
from sklearn.metrics import accuracy_score, confusion_matrix

import tensorflow as tf
from tensorflow.keras.layers import Dense, Input
from tensorflow.keras.models import Model
from tensorflow.keras import optimizers

from matplotlib import pyplot as plt
import seaborn as sns
sns.set()
from damavand.damavand.datasets.downloaders import read_addresses, ZipDatasetDownloader
from damavand.damavand.datasets.digestors import SEU
from damavand.damavand.signal_processing import *
from damavand.damavand.utils import *

import pandas as pd
from scipy.stats import skew, kurtosis

from sklearn.preprocessing import LabelEncoder, LabelBinarizer, MinMaxScaler
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier
from sklearn.metrics import accuracy_score, confusion_matrix

import tensorflow as tf
from tensorflow.keras.layers import Dense, Input
from tensorflow.keras.models import Model
from tensorflow.keras import optimizers

from matplotlib import pyplot as plt
import seaborn as sns
sns.set()

Instantiating a downloader object¶

In [4]:

Copied!

addresses = read_addresses()
downloader = ZipDatasetDownloader(addresses['SEU'])
downloader.download_extract('SEU.zip', 'SEU/')
addresses = read_addresses()
downloader = ZipDatasetDownloader(addresses['SEU'])
downloader.download_extract('SEU.zip', 'SEU/')

Instantiating a digestor object¶

In [5]:

Copied!





seu = SEU('SEU/')
mining_params = {
    'win_len': 10000,
    'hop_len': 10000
}
seu.mine(mining_params)
seu = SEU('SEU/')
mining_params = {
    'win_len': 10000,
    'hop_len': 10000
}
seu.mine(mining_params)

Mining:  Chipped_20_0.csv
Mining:  Miss_20_0.csv
Mining:  Root_30_2.csv
Mining:  Surface_20_0.csv
Mining:  Health_20_0.csv
Mining:  Root_20_0.csv
Mining:  Miss_30_2.csv
Mining:  Health_30_2.csv
Mining:  Chipped_30_2.csv
Mining:  Surface_30_2.csv
Mining:  ball_20_0.csv
Mining:  comb_20_0.csv
Mining:  ball_30_2.csv
Mining:  inner_20_0.csv
Mining:  health_20_0.csv
Mining:  comb_30_2.csv
Mining:  outer_30_2.csv
Mining:  health_30_2.csv
Mining:  inner_30_2.csv
Mining:  outer_20_0.csv

Aggregating data over the second channel¶

In [6]:

Copied!

df = pd.concat(seu.data[1]).reset_index(drop = True)
df
df = pd.concat(seu.data[1]).reset_index(drop = True)
df

Out[6]:

	0	1	2	3	4	5	6	7	8	9	...	9993	9994	9995	9996	9997	9998	9999	test_bed	state	rot_speed
0	0.000503	0.002112	0.003276	0.002187	0.000864	0.003220	-0.000612	0.000082	0.000796	0.003134	...	-0.001155	0.000290	0.000763	-0.000211	0.002804	0.003674	0.000706	Gear_set	Chipped	20
1	0.003633	-0.000229	0.000552	-0.000004	0.002151	0.003697	0.001304	0.004830	0.004209	0.003994	...	0.000752	-0.002675	0.002458	0.002286	-0.001379	0.002416	0.005218	Gear_set	Chipped	20
2	0.002939	-0.000224	0.000743	0.000638	0.000941	0.004123	-0.001305	0.001551	0.005114	0.000157	...	0.001551	0.000154	0.001876	0.001844	0.001475	0.001525	0.000941	Gear_set	Chipped	20
3	0.000901	0.001673	0.000249	0.001255	0.001456	0.000186	0.001425	0.000232	-0.000758	0.002090	...	0.001602	0.000980	0.001359	0.007073	0.002937	0.004625	0.004481	Gear_set	Chipped	20
4	0.003072	0.000039	0.000633	0.000852	-0.000666	0.001584	0.000923	0.002730	0.004590	0.000892	...	-0.002054	-0.002071	0.003704	0.000231	0.003415	0.005243	0.000980	Gear_set	Chipped	20
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2075	0.003477	0.004098	-0.000902	0.002244	-0.000417	-0.000520	0.001959	0.001307	-0.002608	0.004534	...	-0.000941	-0.001752	-0.000422	-0.001873	-0.000856	0.001448	0.001341	Bearing_set	outer	20
2076	0.001140	0.001454	0.002049	0.000772	0.001109	-0.001004	-0.000777	0.000978	0.000858	0.001173	...	-0.000085	0.000597	-0.000103	0.000166	0.000979	0.000370	0.000114	Bearing_set	outer	20
2077	0.000862	-0.000015	0.000430	0.000468	0.000056	0.001246	0.001090	0.000552	0.000275	-0.000464	...	-0.001698	0.002050	0.000086	0.000114	0.002718	0.000453	0.000899	Bearing_set	outer	20
2078	0.002031	0.000311	0.000124	-0.002179	-0.001022	-0.000153	-0.001156	-0.000972	-0.000298	-0.000947	...	0.000472	-0.001147	0.001435	-0.000050	0.000645	-0.002612	0.003531	Bearing_set	outer	20
2079	0.000778	-0.002289	0.002877	0.000565	-0.000243	0.002654	0.000502	-0.003823	0.002394	-0.001581	...	0.001876	0.001363	-0.000221	0.002683	0.003662	0.002092	0.002457	Bearing_set	outer	20

2080 rows × 10003 columns

Signals-Metadata declaration¶

In [7]:

Copied!

signals, metadata = df.iloc[:, : - 3], df.iloc[:, - 3 :]
signals
signals, metadata = df.iloc[:, : - 3], df.iloc[:, - 3 :]
signals

Out[7]:

	0	1	2	3	4	5	6	7	8	9	...	9990	9991	9992	9993	9994	9995	9996	9997	9998	9999
0	0.000503	0.002112	0.003276	0.002187	0.000864	0.003220	-0.000612	0.000082	0.000796	0.003134	...	0.001869	0.001483	0.000743	-0.001155	0.000290	0.000763	-0.000211	0.002804	0.003674	0.000706
1	0.003633	-0.000229	0.000552	-0.000004	0.002151	0.003697	0.001304	0.004830	0.004209	0.003994	...	0.002317	0.000151	-0.000625	0.000752	-0.002675	0.002458	0.002286	-0.001379	0.002416	0.005218
2	0.002939	-0.000224	0.000743	0.000638	0.000941	0.004123	-0.001305	0.001551	0.005114	0.000157	...	0.001116	0.000269	0.000259	0.001551	0.000154	0.001876	0.001844	0.001475	0.001525	0.000941
3	0.000901	0.001673	0.000249	0.001255	0.001456	0.000186	0.001425	0.000232	-0.000758	0.002090	...	0.000246	-0.000108	0.001671	0.001602	0.000980	0.001359	0.007073	0.002937	0.004625	0.004481
4	0.003072	0.000039	0.000633	0.000852	-0.000666	0.001584	0.000923	0.002730	0.004590	0.000892	...	0.001267	0.003988	0.002531	-0.002054	-0.002071	0.003704	0.000231	0.003415	0.005243	0.000980
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2075	0.003477	0.004098	-0.000902	0.002244	-0.000417	-0.000520	0.001959	0.001307	-0.002608	0.004534	...	-0.001024	-0.000501	-0.000623	-0.000941	-0.001752	-0.000422	-0.001873	-0.000856	0.001448	0.001341
2076	0.001140	0.001454	0.002049	0.000772	0.001109	-0.001004	-0.000777	0.000978	0.000858	0.001173	...	0.000672	0.000901	0.000944	-0.000085	0.000597	-0.000103	0.000166	0.000979	0.000370	0.000114
2077	0.000862	-0.000015	0.000430	0.000468	0.000056	0.001246	0.001090	0.000552	0.000275	-0.000464	...	0.000809	0.004516	0.002311	-0.001698	0.002050	0.000086	0.000114	0.002718	0.000453	0.000899
2078	0.002031	0.000311	0.000124	-0.002179	-0.001022	-0.000153	-0.001156	-0.000972	-0.000298	-0.000947	...	0.000643	0.000139	0.000671	0.000472	-0.001147	0.001435	-0.000050	0.000645	-0.002612	0.003531
2079	0.000778	-0.002289	0.002877	0.000565	-0.000243	0.002654	0.000502	-0.003823	0.002394	-0.001581	...	0.000547	0.001853	0.001171	0.001876	0.001363	-0.000221	0.002683	0.003662	0.002092	0.002457

2080 rows × 10000 columns

In [8]:

Copied!

metadata
metadata

Out[8]:

	test_bed	state	rot_speed
0	Gear_set	Chipped	20
1	Gear_set	Chipped	20
2	Gear_set	Chipped	20
3	Gear_set	Chipped	20
4	Gear_set	Chipped	20
...	...	...	...
2075	Bearing_set	outer	20
2076	Bearing_set	outer	20
2077	Bearing_set	outer	20
2078	Bearing_set	outer	20
2079	Bearing_set	outer	20

2080 rows × 3 columns

Signal Processing¶

FFT¶

In [9]:

Copied!

window = scipy.signal.windows.hann(signals.shape[1])
freq_filter = scipy.signal.butter(25, [15, 950], 'bandpass', fs = 2000, output='sos')
window = scipy.signal.windows.hann(signals.shape[1])
freq_filter = scipy.signal.butter(25, [15, 950], 'bandpass', fs = 2000, output='sos')

In [10]:

Copied!

signals_fft = fft(signals, freq_filter = freq_filter, window = window)
signals_fft.shape
signals_fft = fft(signals, freq_filter = freq_filter, window = window)
signals_fft.shape

Out[10]:

(2080, 5000)

Statistical Features¶

In [11]:

Copied!

signals_features = feature(signals, {'mean': np.mean, 'var': np.var, 'rms': rms, 'skew': skew, 'kurtosis': kurtosis})
signals_features.shape
signals_features = feature(signals, {'mean': np.mean, 'var': np.var, 'rms': rms, 'skew': skew, 'kurtosis': kurtosis})
signals_features.shape

Out[11]:

(2080, 5)

Health classification task¶

Target declaration¶

In [12]:

Copied!





y = metadata['state']
le = LabelEncoder()
y_encoded = le.fit_transform(y)
np.unique(y_encoded, return_counts = True)
y = metadata['state']
le = LabelEncoder()
y_encoded = le.fit_transform(y)
np.unique(y_encoded, return_counts = True)

Out[12]:

(array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]),
 array([208, 208, 208, 208, 208, 208, 208, 208, 208, 208]))

In [13]:

Copied!

lb = LabelBinarizer()
y_bin = lb.fit_transform(y_encoded)
lb = LabelBinarizer()
y_bin = lb.fit_transform(y_encoded)

Train-test split¶

In [14]:

Copied!

x_fft_train, x_fft_test, x_features_train, x_features_test, y_encoded_train, y_encoded_test, y_bin_train, y_bin_test = train_test_split(signals_fft, signals_features, y_encoded, y_bin, test_size=0.33, random_state=42)
x_fft_train, x_fft_test, x_features_train, x_features_test, y_encoded_train, y_encoded_test, y_bin_train, y_bin_test = train_test_split(signals_fft, signals_features, y_encoded, y_bin, test_size=0.33, random_state=42)

Training¶

Classic machine learning¶

In [15]:

Copied!

DT = DecisionTreeClassifier()
DT.fit(x_features_train, y_encoded_train)
DT = DecisionTreeClassifier()
DT.fit(x_features_train, y_encoded_train)

Out[15]:

DecisionTreeClassifier()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Deep Learning¶

In [16]:

Copied!

def model_creator():

  input = Input(shape = (5000,), name='input1')
  e = Dense(units = 2500, activation='tanh', name = 'HL1')(input)
  predicted_label = Dense(units = 10, activation='softmax', name='label')(e)

  return Model(inputs = input, outputs = predicted_label)
def model_creator():

  input = Input(shape = (5000,), name='input1')
  e = Dense(units = 2500, activation='tanh', name = 'HL1')(input)
  predicted_label = Dense(units = 10, activation='softmax', name='label')(e)

  return Model(inputs = input, outputs = predicted_label)

In [17]:

Copied!

scaler = MinMaxScaler()
x_fft_train_scaled = scaler.fit_transform(x_fft_train)
x_fft_test_scaled = scaler.transform(x_fft_test)
scaler = MinMaxScaler()
x_fft_train_scaled = scaler.fit_transform(x_fft_train)
x_fft_test_scaled = scaler.transform(x_fft_test)

In [18]:

Copied!





lr = 0.0001
ep = 25

model = model_creator()
opt = optimizers.Adam(learning_rate=lr, decay=lr / ep)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['accuracy'])

histories  = model.fit(x_fft_train_scaled, y_bin_train, validation_split=0.25, epochs = ep, batch_size = 250)
lr = 0.0001
ep = 25

model = model_creator()
opt = optimizers.Adam(learning_rate=lr, decay=lr / ep)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['accuracy'])

histories  = model.fit(x_fft_train_scaled, y_bin_train, validation_split=0.25, epochs = ep, batch_size = 250)

Epoch 1/25

/usr/local/lib/python3.10/dist-packages/keras/src/optimizers/base_optimizer.py:33: UserWarning: Argument `decay` is no longer supported and will be ignored.
  warnings.warn(

5/5 ━━━━━━━━━━━━━━━━━━━━ 4s 737ms/step - accuracy: 0.1329 - loss: 2.5406 - val_accuracy: 0.2636 - val_loss: 1.9598
Epoch 2/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 4s 412ms/step - accuracy: 0.2904 - loss: 1.8779 - val_accuracy: 0.6017 - val_loss: 1.4486
Epoch 3/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 405ms/step - accuracy: 0.6108 - loss: 1.3780 - val_accuracy: 0.7593 - val_loss: 1.1201
Epoch 4/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 407ms/step - accuracy: 0.8138 - loss: 1.0299 - val_accuracy: 0.9828 - val_loss: 0.8795
Epoch 5/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 407ms/step - accuracy: 0.9521 - loss: 0.8040 - val_accuracy: 0.9628 - val_loss: 0.7117
Epoch 6/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 3s 647ms/step - accuracy: 0.9742 - loss: 0.6438 - val_accuracy: 0.9857 - val_loss: 0.5797
Epoch 7/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 4s 408ms/step - accuracy: 0.9917 - loss: 0.4939 - val_accuracy: 1.0000 - val_loss: 0.4681
Epoch 8/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 408ms/step - accuracy: 1.0000 - loss: 0.4052 - val_accuracy: 1.0000 - val_loss: 0.4004
Epoch 9/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 407ms/step - accuracy: 1.0000 - loss: 0.3366 - val_accuracy: 1.0000 - val_loss: 0.3459
Epoch 10/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 3s 455ms/step - accuracy: 1.0000 - loss: 0.2738 - val_accuracy: 1.0000 - val_loss: 0.2988
Epoch 11/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 4s 768ms/step - accuracy: 1.0000 - loss: 0.2409 - val_accuracy: 1.0000 - val_loss: 0.2632
Epoch 12/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 3s 412ms/step - accuracy: 1.0000 - loss: 0.2074 - val_accuracy: 1.0000 - val_loss: 0.2260
Epoch 13/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 3s 410ms/step - accuracy: 1.0000 - loss: 0.1778 - val_accuracy: 1.0000 - val_loss: 0.2030
Epoch 14/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 406ms/step - accuracy: 1.0000 - loss: 0.1610 - val_accuracy: 1.0000 - val_loss: 0.1859
Epoch 15/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 433ms/step - accuracy: 1.0000 - loss: 0.1504 - val_accuracy: 1.0000 - val_loss: 0.1671
Epoch 16/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 4s 687ms/step - accuracy: 1.0000 - loss: 0.1291 - val_accuracy: 1.0000 - val_loss: 0.1518
Epoch 17/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 414ms/step - accuracy: 1.0000 - loss: 0.1148 - val_accuracy: 1.0000 - val_loss: 0.1380
Epoch 18/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 3s 451ms/step - accuracy: 1.0000 - loss: 0.1058 - val_accuracy: 1.0000 - val_loss: 0.1291
Epoch 19/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 414ms/step - accuracy: 1.0000 - loss: 0.0931 - val_accuracy: 1.0000 - val_loss: 0.1182
Epoch 20/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 3s 409ms/step - accuracy: 1.0000 - loss: 0.0856 - val_accuracy: 1.0000 - val_loss: 0.1096
Epoch 21/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 4s 696ms/step - accuracy: 1.0000 - loss: 0.0815 - val_accuracy: 1.0000 - val_loss: 0.1015
Epoch 22/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 413ms/step - accuracy: 1.0000 - loss: 0.0769 - val_accuracy: 1.0000 - val_loss: 0.0955
Epoch 23/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 3s 416ms/step - accuracy: 1.0000 - loss: 0.0667 - val_accuracy: 1.0000 - val_loss: 0.0898
Epoch 24/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 421ms/step - accuracy: 1.0000 - loss: 0.0644 - val_accuracy: 1.0000 - val_loss: 0.0832
Epoch 25/25
5/5 ━━━━━━━━━━━━━━━━━━━━ 2s 414ms/step - accuracy: 1.0000 - loss: 0.0593 - val_accuracy: 1.0000 - val_loss: 0.0778

In [19]:

Copied!





fig, axes = plt.subplots(1, 2, figsize=(16,4))
fig.suptitle('Deep Learning Model Training Process')
axes[0].plot(histories.history['loss'], label='Training Loss')
axes[0].plot(histories.history['val_loss'], label='Validation Loss')
axes[0].set_xlabel('Epochs')
axes[0].set_ylabel('Loss')
axes[0].set_title('Loss Vs. Epochs')
axes[0].legend()

axes[1].plot(histories.history['accuracy'], label='Training Accuracy')
axes[1].plot(histories.history['val_accuracy'], label='Validation Accuracy')
axes[1].set_xlabel('Epochs')
axes[1].set_ylabel('Accuracy')
axes[1].set_title('Accuracy Vs. Epochs')
axes[1].legend()
plt.show()
fig, axes = plt.subplots(1, 2, figsize=(16,4))
fig.suptitle('Deep Learning Model Training Process')
axes[0].plot(histories.history['loss'], label='Training Loss')
axes[0].plot(histories.history['val_loss'], label='Validation Loss')
axes[0].set_xlabel('Epochs')
axes[0].set_ylabel('Loss')
axes[0].set_title('Loss Vs. Epochs')
axes[0].legend()

axes[1].plot(histories.history['accuracy'], label='Training Accuracy')
axes[1].plot(histories.history['val_accuracy'], label='Validation Accuracy')
axes[1].set_xlabel('Epochs')
axes[1].set_ylabel('Accuracy')
axes[1].set_title('Accuracy Vs. Epochs')
axes[1].legend()
plt.show()

No description has been provided for this image

Evaluation¶

In [21]:

Copied!





y_train_pred_DT = DT.predict(x_features_train)
y_test_pred_DT = DT.predict(x_features_test)
y_train_pred_DL = np.argmax(model.predict(x_fft_train_scaled), axis = 1)
y_test_pred_DL = np.argmax(model.predict(x_fft_test_scaled), axis = 1)
y_train_pred_DT = DT.predict(x_features_train)
y_test_pred_DT = DT.predict(x_features_test)
y_train_pred_DL = np.argmax(model.predict(x_fft_train_scaled), axis = 1)
y_test_pred_DL = np.argmax(model.predict(x_fft_test_scaled), axis = 1)

44/44 ━━━━━━━━━━━━━━━━━━━━ 1s 27ms/step
22/22 ━━━━━━━━━━━━━━━━━━━━ 1s 44ms/step

Accuracies¶

In [22]:

Copied!





print('Decision tree training accuracy: ', np.round(accuracy_score(y_encoded_train, y_train_pred_DT), 4))
print('Decision tree testing accuracy: ', np.round(accuracy_score(y_encoded_test, y_test_pred_DT), 4))
print('Deep learnign model training accuracy: ', np.round(accuracy_score(y_encoded_train, y_train_pred_DL), 4))
print('Deep learnign model testing accuracy: ', np.round(accuracy_score(y_encoded_test, y_test_pred_DL),  4))
print('Decision tree training accuracy: ', np.round(accuracy_score(y_encoded_train, y_train_pred_DT), 4))
print('Decision tree testing accuracy: ', np.round(accuracy_score(y_encoded_test, y_test_pred_DT), 4))
print('Deep learnign model training accuracy: ', np.round(accuracy_score(y_encoded_train, y_train_pred_DL), 4))
print('Deep learnign model testing accuracy: ', np.round(accuracy_score(y_encoded_test, y_test_pred_DL),  4))

Decision tree training accuracy:  1.0
Decision tree testing accuracy:  0.8195
Deep learnign model training accuracy:  1.0
Deep learnign model testing accuracy:  1.0

Confusion Matrices¶

In [24]:

Copied!

confusion_matrix_DT = confusion_matrix(y_encoded_test, y_test_pred_DT)
confusion_matrix_DL = confusion_matrix(y_encoded_test, y_test_pred_DL)
confusion_matrix_DT = confusion_matrix(y_encoded_test, y_test_pred_DT)
confusion_matrix_DL = confusion_matrix(y_encoded_test, y_test_pred_DL)

In [44]:

Copied!





fig, axes = plt.subplots(1, 2, figsize=(12,4))

sns.heatmap(confusion_matrix_DT, annot=True, square = True, cmap='Blues', fmt='g', cbar = False, ax=axes[0])
axes[0].set_xlabel("Predicted", fontsize=14, labelpad=20)
axes[0].xaxis.set_ticklabels(le.classes_)
axes[0].set_ylabel("Actual", fontsize=14, labelpad=20)
axes[0].yaxis.set_ticklabels(le.classes_)
axes[0].set_title("Decision Tree Confusion Matrix", fontsize=14, pad=20)
axes[0].tick_params(axis='x', rotation=90)
axes[0].tick_params(axis='y', rotation=360)

sns.heatmap(confusion_matrix_DL, annot=True, square = True, cmap='Blues', fmt='g', cbar = False, ax=axes[1])
axes[1].set_xlabel("Predicted", fontsize=14, labelpad=20)
axes[1].xaxis.set_ticklabels(le.classes_)
axes[1].set_ylabel("Actual", fontsize=14, labelpad=20)
axes[1].yaxis.set_ticklabels(le.classes_)
axes[1].set_title("Deep Learning Confusion Matrix", fontsize=14, pad=20)
axes[1].tick_params(axis='x', rotation=90)
axes[1].tick_params(axis='y', rotation=360)

plt.show()
fig, axes = plt.subplots(1, 2, figsize=(12,4))

sns.heatmap(confusion_matrix_DT, annot=True, square = True, cmap='Blues', fmt='g', cbar = False, ax=axes[0])
axes[0].set_xlabel("Predicted", fontsize=14, labelpad=20)
axes[0].xaxis.set_ticklabels(le.classes_)
axes[0].set_ylabel("Actual", fontsize=14, labelpad=20)
axes[0].yaxis.set_ticklabels(le.classes_)
axes[0].set_title("Decision Tree Confusion Matrix", fontsize=14, pad=20)
axes[0].tick_params(axis='x', rotation=90)
axes[0].tick_params(axis='y', rotation=360)

sns.heatmap(confusion_matrix_DL, annot=True, square = True, cmap='Blues', fmt='g', cbar = False, ax=axes[1])
axes[1].set_xlabel("Predicted", fontsize=14, labelpad=20)
axes[1].xaxis.set_ticklabels(le.classes_)
axes[1].set_ylabel("Actual", fontsize=14, labelpad=20)
axes[1].yaxis.set_ticklabels(le.classes_)
axes[1].set_title("Deep Learning Confusion Matrix", fontsize=14, pad=20)
axes[1].tick_params(axis='x', rotation=90)
axes[1].tick_params(axis='y', rotation=360)

plt.show()