DiCE Explanations for Autoencoder model having Time series Data

[iuush]

[gaugup]

@iuush could you provide more description as to what you are looking for?

Regards,

[iuush]

Dear Gaurav,

I am using an Autoencoder model for finding the anomaly in the tabular data. But your package is not working for my model and code.

PFA my code(Notebook).

On Tue, Mar 22, 2022 at 12:07 AM Gaurav Gupta @.***> wrote:

@iuush https://github.com/iuush could you provide more description as to what you are looking for?

Regards,

— Reply to this email directly, view it on GitHub https://github.com/interpretml/DiCE/issues/275#issuecomment-1074273729, or unsubscribe https://github.com/notifications/unsubscribe-auth/AT7CZOY6ASRFBUD727676ZLVBC6VDANCNFSM5QYBOBJA . You are receiving this because you were mentioned.Message ID: @.***>

[iuush]

Dear Gaurav,

I need it urgently. Could you please help me ASAP.

Thanks Ayush +91-9528553052

On Tue, Mar 22, 2022 at 11:53 AM Ayush @.***> wrote:

Dear Gaurav,

I am using an Autoencoder model for finding the anomaly in the tabular data. But your package is not working for my model and code.

PFA my code(Notebook).

On Tue, Mar 22, 2022 at 12:07 AM Gaurav Gupta @.***> wrote:

@iuush https://github.com/iuush could you provide more description as to what you are looking for?

Regards,

— Reply to this email directly, view it on GitHub https://github.com/interpretml/DiCE/issues/275#issuecomment-1074273729, or unsubscribe https://github.com/notifications/unsubscribe-auth/AT7CZOY6ASRFBUD727676ZLVBC6VDANCNFSM5QYBOBJA . You are receiving this because you were mentioned.Message ID: @.***>

[iuush]

Dear Gaurav,

I am using an Autoencoder model for finding the anomaly in the tabular data. But your package is not working for my model and code.

PFA my Model()

On Tue, Mar 22, 2022 at 2:29 PM Ayush @.***> wrote:

Dear Gaurav,

I need it urgently. Could you please help me ASAP.

Thanks Ayush +91-9528553052

On Tue, Mar 22, 2022 at 11:53 AM Ayush @.***> wrote:

Dear Gaurav,

I am using an Autoencoder model for finding the anomaly in the tabular data. But your package is not working for my model and code.

PFA my code(Notebook).

On Tue, Mar 22, 2022 at 12:07 AM Gaurav Gupta @.***> wrote:

@iuush https://github.com/iuush could you provide more description as to what you are looking for?

Regards,

— Reply to this email directly, view it on GitHub https://github.com/interpretml/DiCE/issues/275#issuecomment-1074273729, or unsubscribe https://github.com/notifications/unsubscribe-auth/AT7CZOY6ASRFBUD727676ZLVBC6VDANCNFSM5QYBOBJA . You are receiving this because you were mentioned.Message ID: @.***>

[iuush]

Dear Gaurav,

I need it urgently. Could you please help me ASAP.

Thanks Ayush +91-9528553052

On Tue, Mar 22, 2022 at 4:35 PM Ayush @.***> wrote:

Dear Gaurav,

I am using an Autoencoder model for finding the anomaly in the tabular data. But your package is not working for my model and code.

PFA my Model()

On Tue, Mar 22, 2022 at 2:29 PM Ayush @.***> wrote:

Dear Gaurav,

I need it urgently. Could you please help me ASAP.

Thanks Ayush +91-9528553052

On Tue, Mar 22, 2022 at 11:53 AM Ayush @.***> wrote:

Dear Gaurav,

I am using an Autoencoder model for finding the anomaly in the tabular data. But your package is not working for my model and code.

PFA my code(Notebook).

On Tue, Mar 22, 2022 at 12:07 AM Gaurav Gupta @.***> wrote:

@iuush https://github.com/iuush could you provide more description as to what you are looking for?

Regards,

— Reply to this email directly, view it on GitHub https://github.com/interpretml/DiCE/issues/275#issuecomment-1074273729, or unsubscribe https://github.com/notifications/unsubscribe-auth/AT7CZOY6ASRFBUD727676ZLVBC6VDANCNFSM5QYBOBJA . You are receiving this because you were mentioned.Message ID: @.***>

[amit-sharma]

@iuush I looked through your messages above and did not find the attached notebook or model. Can you reshare it?

We need a minimum reproducible code to understand the issue.

[gaugup]

@iuush, yes hwat @amit-sharma said we need some minimum reproducible code to see what problems you are running into. Even I don't see any notebook attached in this issue.

Regards,

[iuush]

Dear Interpretml/Dice team,

This model is not working for a LSTM autoencoder model or any autoencoder model.

My autoencoder model is:

create a model by subclassing Model class in tensorflow

class AutoEncoder(Model): """ Parameters

output_units: int Number of output units

code_size: int Number of units in bottle neck """

def init(self, output_units, code_size=4): super().init() self.encoder = Sequential([ Dense(64, activation='relu'), Dropout(0.1), Dense(32, activation='relu'), Dropout(0.1), Dense(16, activation='relu'), Dropout(0.1), Dense(code_size, activation='relu') ]) self.decoder = Sequential([ Dense(16, activation='relu'), Dropout(0.1), Dense(32, activation='relu'), Dropout(0.1), Dense(64, activation='relu'), Dropout(0.1), Dense(output_units, activation='sigmoid') ])

def call(self, inputs): encoded = self.encoder(inputs) decoded = self.decoder(encoded) return decoded

def predict_proba(self, inputs): encoded = self.encoder(inputs) decoded = self.decoder(encoded) return 0.1

model = AutoEncoder(output_units=X_train_s.shape[1])

model.compile(loss='mse',optimizer='adam')

X_pred = model.predict(df_s)

regards Ayush

On Tue, Mar 22, 2022 at 7:08 PM Gaurav Gupta @.***> wrote:

@iuush https://github.com/iuush, yes hwat @amit-sharma https://github.com/amit-sharma said we need some minimum reproducible code to see what problems you are running into. Even I don't see any notebook attached in this issue.

Regards,

— Reply to this email directly, view it on GitHub https://github.com/interpretml/DiCE/issues/275#issuecomment-1075191303, or unsubscribe https://github.com/notifications/unsubscribe-auth/AT7CZO2LYPXBJYFSFROEF43VBHENRANCNFSM5QYBOBJA . You are receiving this because you were mentioned.Message ID: @.***>

[gaugup]

@iuush how did you setup the dice-ml explainer to generate counterfactuals? Can you share the dice-ml code that is giving you errors?

Regards,

[iuush]

Dear Gaurav,

Could you please share your contact number. It would be very helpful.

Thanks Ayush

On Wed, Mar 30, 2022 at 4:56 PM Ayush @.***> wrote:

Dear Interpretml/Dice team,

This model is not working for a LSTM autoencoder model or any autoencoder model.

My autoencoder model is:

create a model by subclassing Model class in tensorflow

class AutoEncoder(Model): """ Parameters

output_units: int Number of output units

code_size: int Number of units in bottle neck """

def init(self, output_units, code_size=4): super().init() self.encoder = Sequential([ Dense(64, activation='relu'), Dropout(0.1), Dense(32, activation='relu'), Dropout(0.1), Dense(16, activation='relu'), Dropout(0.1), Dense(code_size, activation='relu') ]) self.decoder = Sequential([ Dense(16, activation='relu'), Dropout(0.1), Dense(32, activation='relu'), Dropout(0.1), Dense(64, activation='relu'), Dropout(0.1), Dense(output_units, activation='sigmoid') ])

def call(self, inputs): encoded = self.encoder(inputs) decoded = self.decoder(encoded) return decoded

def predict_proba(self, inputs): encoded = self.encoder(inputs) decoded = self.decoder(encoded) return 0.1

model = AutoEncoder(output_units=X_train_s.shape[1])

model.compile(loss='mse',optimizer='adam')

X_pred = model.predict(df_s)

regards Ayush

On Tue, Mar 22, 2022 at 7:08 PM Gaurav Gupta @.***> wrote:

@iuush https://github.com/iuush, yes hwat @amit-sharma https://github.com/amit-sharma said we need some minimum reproducible code to see what problems you are running into. Even I don't see any notebook attached in this issue.

Regards,

— Reply to this email directly, view it on GitHub https://github.com/interpretml/DiCE/issues/275#issuecomment-1075191303, or unsubscribe https://github.com/notifications/unsubscribe-auth/AT7CZO2LYPXBJYFSFROEF43VBHENRANCNFSM5QYBOBJA . You are receiving this because you were mentioned.Message ID: @.***>