* 겪고 있는 문제 상황을 최대한 자세하게 작성해주세요.

* 문제 해결을 위해 어떤 시도를 해보았는지 구체적으로 함께 알려주세요.

해당 사진에서 처럼 loss와 val_loss의 숫자가 처음부터 엄청 크게 나오며, 줄어들어도 많이 줄어들지가 않은데

1. 왜 이렇게 나오는지에 대한 이유와
2. loss함수에 MSE가 아닌 RMSE, MAPE, MAE등의 성능지표를 사용해도 되는지 궁금합니다.
3. 또한 SGD는 어느정도 잘나오는데 Adam으로 출력하였을 때 그래프가 왜 아래와 같이 예상과 다르게 나오는지 궁금합니다.

작성한 코드 및 에러 메세지

model = Sequential([

  Dense(1)

])

model.compile(loss='mean_squared_error', optimizer=SGD(lr=0.1))

model.fit(

  x_train,

  y_train,

  validation_data=(x_val, y_val),

  epochs=100

)

결과

Epoch 1/100
1/1 [==============================] - 0s 447ms/step - loss: 7459466752.0000 - val_loss: 174682208.0000
Epoch 2/100
1/1 [==============================] - 0s 47ms/step - loss: 230217744.0000 - val_loss: 183207392.0000
Epoch 3/100
1/1 [==============================] - 0s 35ms/step - loss: 220419280.0000 - val_loss: 182469008.0000
Epoch 4/100
1/1 [==============================] - 0s 36ms/step - loss: 218538576.0000 - val_loss: 181061152.0000
Epoch 5/100
1/1 [==============================] - 0s 38ms/step - loss: 216685216.0000 - val_loss: 179644784.0000
Epoch 6/100
1/1 [==============================] - 0s 41ms/step - loss: 214850432.0000 - val_loss: 178241936.0000
Epoch 7/100
1/1 [==============================] - 0s 42ms/step - loss: 213034112.0000 - val_loss: 176853136.0000
Epoch 8/100
1/1 [==============================] - 0s 39ms/step - loss: 211236000.0000 - val_loss: 175478352.0000
Epoch 9/100
1/1 [==============================] - 0s 42ms/step - loss: 209455920.0000 - val_loss: 174117344.0000
Epoch 10/100
1/1 [==============================] - 0s 54ms/step - loss: 207693648.0000 - val_loss: 172770080.0000
Epoch 11/100
1/1 [==============================] - 0s 56ms/step - loss: 205949136.0000 - val_loss: 171436208.0000
Epoch 12/100
1/1 [==============================] - 0s 63ms/step - loss: 204221936.0000 - val_loss: 170115920.0000
Epoch 13/100
1/1 [==============================] - 0s 58ms/step - loss: 202512288.0000 - val_loss: 168808720.0000
Epoch 14/100
1/1 [==============================] - 0s 64ms/step - loss: 200819600.0000 - val_loss: 167514864.0000
Epoch 15/100
1/1 [==============================] - 0s 78ms/step - loss: 199144064.0000 - val_loss: 166233744.0000
Epoch 16/100
1/1 [==============================] - 0s 52ms/step - loss: 197485168.0000 - val_loss: 164965712.0000
Epoch 17/100
1/1 [==============================] - 0s 58ms/step - loss: 195843040.0000 - val_loss: 163710352.0000
Epoch 18/100
1/1 [==============================] - 0s 67ms/step - loss: 194217264.0000 - val_loss: 162467712.0000
Epoch 19/100
1/1 [==============================] - 0s 76ms/step - loss: 192607920.0000 - val_loss: 161237344.0000
Epoch 20/100
1/1 [==============================] - 0s 49ms/step - loss: 191014608.0000 - val_loss: 160019536.0000
Epoch 21/100
1/1 [==============================] - 0s 48ms/step - loss: 189437296.0000 - val_loss: 158813840.0000
Epoch 22/100
1/1 [==============================] - 0s 49ms/step - loss: 187875840.0000 - val_loss: 157620320.0000
Epoch 23/100
1/1 [==============================] - 0s 65ms/step - loss: 186330016.0000 - val_loss: 156438784.0000
Epoch 24/100
1/1 [==============================] - 0s 65ms/step - loss: 184799728.0000 - val_loss: 155269168.0000
Epoch 25/100
1/1 [==============================] - 0s 72ms/step - loss: 183284768.0000 - val_loss: 154111264.0000
Epoch 26/100
1/1 [==============================] - 0s 50ms/step - loss: 181784976.0000 - val_loss: 152964976.0000
Epoch 27/100
1/1 [==============================] - 0s 50ms/step - loss: 180300288.0000 - val_loss: 151830160.0000
Epoch 28/100
1/1 [==============================] - 0s 77ms/step - loss: 178830432.0000 - val_loss: 150706848.0000
Epoch 29/100
1/1 [==============================] - 0s 51ms/step - loss: 177375344.0000 - val_loss: 149594880.0000
Epoch 30/100
1/1 [==============================] - 0s 78ms/step - loss: 175934880.0000 - val_loss: 148493904.0000
Epoch 31/100
1/1 [==============================] - 0s 48ms/step - loss: 174508768.0000 - val_loss: 147404176.0000
Epoch 32/100
1/1 [==============================] - 0s 48ms/step - loss: 173097056.0000 - val_loss: 146325232.0000
Epoch 33/100
1/1 [==============================] - 0s 48ms/step - loss: 171699456.0000 - val_loss: 145257296.0000
Epoch 34/100
1/1 [==============================] - 0s 68ms/step - loss: 170315888.0000 - val_loss: 144200048.0000
Epoch 35/100
1/1 [==============================] - 0s 58ms/step - loss: 168946176.0000 - val_loss: 143153360.0000
Epoch 36/100
1/1 [==============================] - 0s 53ms/step - loss: 167590176.0000 - val_loss: 142117232.0000
Epoch 37/100
1/1 [==============================] - 0s 52ms/step - loss: 166247856.0000 - val_loss: 141091616.0000
Epoch 38/100
1/1 [==============================] - 0s 78ms/step - loss: 164918960.0000 - val_loss: 140076112.0000
Epoch 39/100
1/1 [==============================] - 0s 50ms/step - loss: 163603376.0000 - val_loss: 139070960.0000
Epoch 40/100
1/1 [==============================] - 0s 68ms/step - loss: 162301008.0000 - val_loss: 138075936.0000
Epoch 41/100
1/1 [==============================] - 0s 73ms/step - loss: 161011728.0000 - val_loss: 137090816.0000
Epoch 42/100
1/1 [==============================] - 0s 82ms/step - loss: 159735328.0000 - val_loss: 136115568.0000
Epoch 43/100
1/1 [==============================] - 0s 71ms/step - loss: 158471728.0000 - val_loss: 135150288.0000
Epoch 44/100
1/1 [==============================] - 0s 93ms/step - loss: 157220848.0000 - val_loss: 134194488.0000
Epoch 45/100
1/1 [==============================] - 0s 68ms/step - loss: 155982464.0000 - val_loss: 133248472.0000
Epoch 46/100
1/1 [==============================] - 0s 94ms/step - loss: 154756512.0000 - val_loss: 132311880.0000
Epoch 47/100
1/1 [==============================] - 0s 93ms/step - loss: 153542896.0000 - val_loss: 131384808.0000
Epoch 48/100
1/1 [==============================] - 0s 81ms/step - loss: 152341408.0000 - val_loss: 130466944.0000
Epoch 49/100
1/1 [==============================] - 0s 71ms/step - loss: 151151952.0000 - val_loss: 129558392.0000
Epoch 50/100
1/1 [==============================] - 0s 93ms/step - loss: 149974512.0000 - val_loss: 128658872.0000
Epoch 51/100
1/1 [==============================] - 0s 44ms/step - loss: 148808800.0000 - val_loss: 127768560.0000
Epoch 52/100
1/1 [==============================] - 0s 45ms/step - loss: 147654800.0000 - val_loss: 126887016.0000
Epoch 53/100
1/1 [==============================] - 0s 63ms/step - loss: 146512368.0000 - val_loss: 126014360.0000
Epoch 54/100
1/1 [==============================] - 0s 46ms/step - loss: 145381360.0000 - val_loss: 125150584.0000
Epoch 55/100
1/1 [==============================] - 0s 64ms/step - loss: 144261760.0000 - val_loss: 124295496.0000
Epoch 56/100
1/1 [==============================] - 0s 52ms/step - loss: 143153408.0000 - val_loss: 123448904.0000
Epoch 57/100
1/1 [==============================] - 0s 45ms/step - loss: 142056112.0000 - val_loss: 122610904.0000
Epoch 58/100
1/1 [==============================] - 0s 53ms/step - loss: 140969824.0000 - val_loss: 121781240.0000
Epoch 59/100
1/1 [==============================] - 0s 64ms/step - loss: 139894432.0000 - val_loss: 120960008.0000
Epoch 60/100
1/1 [==============================] - 0s 50ms/step - loss: 138829888.0000 - val_loss: 120147008.0000
Epoch 61/100
1/1 [==============================] - 0s 48ms/step - loss: 137775952.0000 - val_loss: 119342152.0000
Epoch 62/100
1/1 [==============================] - 0s 51ms/step - loss: 136732592.0000 - val_loss: 118545488.0000
Epoch 63/100
1/1 [==============================] - 0s 43ms/step - loss: 135699744.0000 - val_loss: 117756744.0000
Epoch 64/100
1/1 [==============================] - 0s 45ms/step - loss: 134677216.0000 - val_loss: 116975920.0000
Epoch 65/100
1/1 [==============================] - 0s 46ms/step - loss: 133664952.0000 - val_loss: 116202840.0000
Epoch 66/100
1/1 [==============================] - 0s 46ms/step - loss: 132662784.0000 - val_loss: 115437808.0000
Epoch 67/100
1/1 [==============================] - 0s 50ms/step - loss: 131670768.0000 - val_loss: 114680296.0000
Epoch 68/100
1/1 [==============================] - 0s 53ms/step - loss: 130688648.0000 - val_loss: 113930416.0000
Epoch 69/100
1/1 [==============================] - 0s 50ms/step - loss: 129716400.0000 - val_loss: 113188104.0000
Epoch 70/100
1/1 [==============================] - 0s 49ms/step - loss: 128753864.0000 - val_loss: 112453272.0000
Epoch 71/100
1/1 [==============================] - 0s 62ms/step - loss: 127801016.0000 - val_loss: 111725744.0000
Epoch 72/100
1/1 [==============================] - 0s 50ms/step - loss: 126857712.0000 - val_loss: 111005560.0000
Epoch 73/100
1/1 [==============================] - 0s 47ms/step - loss: 125923840.0000 - val_loss: 110292720.0000
Epoch 74/100
1/1 [==============================] - 0s 42ms/step - loss: 124999384.0000 - val_loss: 109586968.0000
Epoch 75/100
1/1 [==============================] - 0s 49ms/step - loss: 124084168.0000 - val_loss: 108888320.0000
Epoch 76/100
1/1 [==============================] - 0s 46ms/step - loss: 123178168.0000 - val_loss: 108196664.0000
Epoch 77/100
1/1 [==============================] - 0s 56ms/step - loss: 122281216.0000 - val_loss: 107512016.0000
Epoch 78/100
1/1 [==============================] - 0s 65ms/step - loss: 121393272.0000 - val_loss: 106834216.0000
Epoch 79/100
1/1 [==============================] - 0s 66ms/step - loss: 120514264.0000 - val_loss: 106163312.0000
Epoch 80/100
1/1 [==============================] - 0s 68ms/step - loss: 119644032.0000 - val_loss: 105499088.0000
Epoch 81/100
1/1 [==============================] - 0s 54ms/step - loss: 118782536.0000 - val_loss: 104841480.0000
Epoch 82/100
1/1 [==============================] - 0s 52ms/step - loss: 117929656.0000 - val_loss: 104190680.0000
Epoch 83/100
1/1 [==============================] - 0s 53ms/step - loss: 117085400.0000 - val_loss: 103546232.0000
Epoch 84/100
1/1 [==============================] - 0s 50ms/step - loss: 116249520.0000 - val_loss: 102908336.0000
Epoch 85/100
1/1 [==============================] - 0s 49ms/step - loss: 115422088.0000 - val_loss: 102276760.0000
Epoch 86/100
1/1 [==============================] - 0s 91ms/step - loss: 114602896.0000 - val_loss: 101651712.0000
Epoch 87/100
1/1 [==============================] - 0s 45ms/step - loss: 113791992.0000 - val_loss: 101032904.0000
Epoch 88/100
1/1 [==============================] - 0s 73ms/step - loss: 112989160.0000 - val_loss: 100420312.0000
Epoch 89/100
1/1 [==============================] - 0s 41ms/step - loss: 112194472.0000 - val_loss: 99813904.0000
Epoch 90/100
1/1 [==============================] - 0s 38ms/step - loss: 111407672.0000 - val_loss: 99213448.0000
Epoch 91/100
1/1 [==============================] - 0s 38ms/step - loss: 110628752.0000 - val_loss: 98619152.0000
Epoch 92/100
1/1 [==============================] - 0s 35ms/step - loss: 109857680.0000 - val_loss: 98030832.0000
Epoch 93/100
1/1 [==============================] - 0s 41ms/step - loss: 109094344.0000 - val_loss: 97448448.0000
Epoch 94/100
1/1 [==============================] - 0s 43ms/step - loss: 108338648.0000 - val_loss: 96871896.0000
Epoch 95/100
1/1 [==============================] - 0s 37ms/step - loss: 107590536.0000 - val_loss: 96301168.0000
Epoch 96/100
1/1 [==============================] - 0s 39ms/step - loss: 106849944.0000 - val_loss: 95736144.0000
Epoch 97/100
1/1 [==============================] - 0s 58ms/step - loss: 106116792.0000 - val_loss: 95176776.0000
Epoch 98/100
1/1 [==============================] - 0s 40ms/step - loss: 105390976.0000 - val_loss: 94623088.0000
Epoch 99/100
1/1 [==============================] - 0s 35ms/step - loss: 104672400.0000 - val_loss: 94075048.0000
Epoch 100/100
1/1 [==============================] - 0s 36ms/step - loss: 103961088.0000 - val_loss: 93532280.0000

<keras.callbacks.History at 0x7a0404286a10>