5) Design a neural Network for classifying news wires (Multi class classification) using Reutersdataset. Code: from keras.datasets import reuters (tr, tl), (te, tel) = reuters.load_data(num_words=10000) d= reuters.get_word_index() rd = dict([(y, x) for (x, y) in d.items()]) me= ' '.join([rd.get(i - 3, '?') for i in tr[0]]) print(me) import numpy as np trm = np.zeros((len(tr), 10000)) for i ,x in enumerate(tr): trm[i,x]=1. tem = np.zeros((len(te),10000)) fori ,x in enumerate(te): tem[i,x]=1. from keras.utils import to_categoricaltl=to_categorical(tl) tel=to_categorical(tel) tx=trm[:1000] ty=tl[:1000] tex=tem[1000:] tey=tel[1000:] from keras.models import Sequential from keras.layers import Dense n=Sequential() l1=Dense(64,activation='relu',input_shape=(10000,)) l2=Dense(64,activation='relu') l3=Dense(46,activation='softmax') n.add(l1) n.add(l2) n.add(l3) n.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['accuracy']) n.fit(tx,ty,epochs=10) yp=n.predict(tex) print("preicted","actual") print(tey[0],np.argmax(yp[0])) Output: ? ? ? said as a result of its december acquisition of space co it expects earnings per share in 1987 of 1 15 to 1 30 dlrs per share up from 70 cts in 1986 the company said pretax net should rise to nine to 10 mlndlrs from six mlndlrs in 1986 and rental operation revenues to 19 to 22 mlndlrs from 12 5 mlndlrs it said cash flow per share this year should be 2 50 to three dlrsreuter 3 Epoch 1/10 32/32 [==============================] - 2s 17ms/step - loss: 3.0454 - accuracy: 0.4500 Epoch 2/10 32/32 [==============================] - 1s 16ms/step - loss: 1.5415 - accuracy: 0.6320 Epoch 3/10 32/32 [==============================] - 0s 15ms/step - loss: 0.9180 - accuracy: 0.8000 Epoch 4/10 32/32 [==============================] - 0s 15ms/step - loss: 0.5608 - accuracy: 0.8790 Epoch 5/10 32/32 [==============================] - 0s 16ms/step - loss: 0.3407 - accuracy: 0.9350 Epoch 6/10 32/32 [==============================] - 1s 16ms/step - loss: 0.2055 - accuracy: 0.9730 Epoch 7/10 32/32 [==============================] - 0s 16ms/step - loss: 0.1249 - accuracy: 0.9850 Epoch 8/10 32/32 [==============================] - 1s 16ms/step - loss: 0.0775 - accuracy: 0.9880 Epoch 9/10 32/32 [==============================] - 0s 15ms/step - loss: 0.0524 - accuracy: 0.9960 Epoch 10/10 32/32 [==============================] - 0s 15ms/step - loss: 0.0388 - accuracy: 0.9940 39/39 [==============================] - 0s 5ms/step preicted actual [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.] 13