improvement of ai model

app/cron_ai_score.py

@@ -317,7 +317,7 @@ async def process_symbol(ticker, con, start_date, end_date):
         split_size = int(len(df) * (1-test_size))
         test_data = df.iloc[split_size:]
         #selected_features = [col for col in df.columns if col not in ['date','price','Target']]
-        best_features = ['freeCashFlowYield', 'cci', 'daily_return', 'cashAndCashEquivalents_to_cashAndShortTermInvestments', 'longTermDebt_to_totalLiabilitiesAndStockholdersEquity', 'longTermDebt_to_totalAssets', 'totalStockholdersEquity_to_totalLiabilitiesAndStockholdersEquity', 'totalStockholdersEquity_to_totalAssets']
+        best_features = ['williams', 'stoch', 'fdi', 'revenue_to_cashAndCashEquivalents', 'revenue_to_cashAndShortTermInvestments', 'costOfRevenue_to_cashAndCashEquivalents', 'costOfRevenue_to_cashAndShortTermInvestments', 'ebitda_to_cashAndShortTermInvestments', 'incomeTaxExpense_to_cashAndCashEquivalents', 'incomeTaxExpense_to_cashAndShortTermInvestments', 'capitalExpenditure_to_cashAndCashEquivalents', 'capitalExpenditure_to_cashAndShortTermInvestments', 'totalCurrentLiabilities_to_cashAndShortTermInvestments', 'netDebt_to_cashAndShortTermInvestments', 'inventory_to_cashAndShortTermInvestments']
         print(f"For the Ticker: {ticker}")
         data = predictor.evaluate_model(test_data[best_features], test_data['Target'])
 
@@ -391,7 +391,7 @@ async def train_process(tickers, con):
     predictor = ScorePredictor()
     #print(selected_features)
     selected_features = [col for col in df_train if col not in ['price','date','Target']]
-    best_features = predictor.feature_selection(df_train[selected_features], df_train['Target'],k=8)
+    best_features = predictor.feature_selection(df_train[selected_features], df_train['Target'],k=15)
     print(best_features)
     predictor.train_model(df_train[best_features], df_train['Target'])
     predictor.evaluate_model(df_test[best_features], df_test['Target'])
@@ -406,7 +406,7 @@ async def run():
     
     if train_mode:
         #Train first model
-        cursor.execute("SELECT DISTINCT symbol FROM stocks WHERE marketCap >= 50E9 AND symbol NOT LIKE '%.%'")
+        cursor.execute("SELECT DISTINCT symbol FROM stocks WHERE marketCap >= 500E9 AND symbol NOT LIKE '%.%'")
         stock_symbols = [row[0] for row in cursor.fetchall()]
         print('Number of Stocks')
         print(len(stock_symbols))

app/ml_models/score_model.py

@@ -38,24 +38,24 @@ class ScorePredictor:
         clear_session()
         
         # Input layer
-        inputs = Input(shape=(8,))
+        inputs = Input(shape=(15,))
         
         # First dense layer
-        x = Dense(512, activation='relu')(inputs)
+        x = Dense(64, activation='leaky_relu')(inputs)
-        x = Dropout(0.5)(x)
+        x = Dropout(0.3)(x)
         x = BatchNormalization()(x)
         
         # Additional dense layers
-        for units in [256,128]:
+        for units in [64,32]:
-            x = Dense(units, activation='relu')(x)
+            x = Dense(units, activation='leaky_relu')(x)
-            x = Dropout(0.5)(x)
+            x = Dropout(0.3)(x)
             x = BatchNormalization()(x)
         
         # Reshape for attention mechanism
-        x = Reshape((128, 1))(x)
+        x = Reshape((32, 1))(x)
         
         # Attention mechanism
-        attention = Dense(128, activation='relu')(x)
+        attention = Dense(32, activation='leaky_relu')(x)
         attention = Dense(1, activation='softmax')(attention)
         
         # Apply attention
@@ -71,7 +71,7 @@ class ScorePredictor:
         model = Model(inputs=inputs, outputs=outputs)
         
         # Optimizer with a lower learning rate
-        optimizer = Adam(learning_rate=0.001, clipnorm=1.0)
+        optimizer = Adam(learning_rate=0.01, clipnorm=1.0)
         
         # Compile the model
         model.compile(optimizer=optimizer, loss='sparse_categorical_crossentropy', metrics=['accuracy'])
@@ -95,8 +95,8 @@ class ScorePredictor:
         checkpoint = ModelCheckpoint('ml_models/weights/ai-score/weights.keras', 
                                       save_best_only=True, save_freq = 1,
                                       monitor='val_loss', mode='min')
-        early_stopping = EarlyStopping(monitor='val_loss', patience=50, restore_best_weights=True)
+        early_stopping = EarlyStopping(monitor='val_loss', patience=20, restore_best_weights=True)
-        reduce_lr = ReduceLROnPlateau(monitor='val_loss', factor=0.5, patience=30, min_lr=0.001)
+        reduce_lr = ReduceLROnPlateau(monitor='val_loss', factor=0.5, patience=10, min_lr=0.001)
 
         self.model.fit(X_train, y_train, epochs=100_000, batch_size=32, 
                        validation_split=0.1, callbacks=[checkpoint, early_stopping, reduce_lr])
@@ -112,7 +112,7 @@ class ScorePredictor:
         
         # Get the model's predictions
         test_predictions = self.model.predict(X_test)
-        #print(test_predictions)
+        print(test_predictions)
 
         # Extract the probabilities for class 1 (index 1 in the softmax output)
         class_1_probabilities = test_predictions[:, 1]