Device component management using deep learning techniques

1. A computer-implemented method comprising:

obtaining telemetry data from one or more enterprise devices;

determining, for each of the one or more enterprise devices, values for multiple device attributes by processing the obtained telemetry data;

generating, for each of the one or more enterprise devices, at least one prediction related to lifecycle information of at least one device component by processing the determined attribute values using one or more deep learning techniques, wherein the one or more deep learning techniques comprise at least one neural network containing at least one transition matrix between two or more hidden states of the at least one neural network, wherein each of the two or more hidden states comprises one or more probability distributions associated with one or more input layers of the at least one neural network; and

performing one or more automated actions based at least in part on the at least one generated prediction;

wherein the method is performed by at least one processing device comprising a processor coupled to a memory.

2. The computer-implemented method of claim 1 , wherein performing the one or more automated actions comprises automatically updating a device component replacement policy.

3. The computer-implemented method of claim 1 , wherein performing the one or more automated actions comprises automatically implementing a device component replacement policy in accordance with the at least one generated prediction.

4. The computer-implemented method of claim 1 , further comprising:

training the one or more deep learning techniques using the at least one generated prediction.

5. The computer-implemented method of claim 1 , wherein the at least one neural network comprises a variable number of hidden layers.

6. The computer-implemented method of claim 1 , wherein the at least one transition matrix between two or more hidden states of the at least one neural network comprises at least one transition matrix between two or more hidden metastable states, wherein each of the two or more hidden metastable states comprises a probability distribution of visiting one or more discrete microstates contained in the one or more input layers of the at least one neural network.

7. The computer-implemented method of claim 1 , wherein the one or more deep learning techniques comprise at least one deep learning technique based at least in part on a hybrid hidden Markov model.

8. The computer-implemented method of claim 1 , wherein the one or more deep learning techniques comprise at least one deep learning technique based at least in part on a doubly stochastic model.

9. The computer-implemented method of claim 1 , wherein the one or more deep learning techniques comprise at least one deep learning technique based at least in part on a hybrid hidden Markov model integrated with a doubly stochastic model.

10. The computer-implemented method of claim 1 , wherein the multiple device attributes comprise two or more of timestamp information, information pertaining to a number of hours the device is in operation, information pertaining to operational speed, information pertaining to counts of relocating, information pertaining to types of controllers, information pertaining to logical block size, information pertaining to rebuild rate, information pertaining to patrol read rate, information pertaining to cache memory size, and information pertaining to feedback failure rates.

11. The computer-implemented method of claim 1 , wherein the one or more enterprise devices comprise one or more storage devices.

12. A non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device:

to obtain telemetry data from one or more enterprise devices;

to determine, for each of the one or more enterprise devices, values for multiple device attributes by processing the obtained telemetry data;

to generate, for each of the one or more enterprise devices, at least one prediction related to lifecycle information of at least one device component by processing the determined attribute values using one or more deep learning techniques, wherein the one or more deep learning techniques comprise at least one neural network containing at least one transition matrix between two or more hidden states of the at least one neural network, wherein each of the two or more hidden states comprises one or more probability distributions associated with one or more input layers of the at least one neural network; and

to perform one or more automated actions based at least in part on the at least one generated prediction.

13. The non-transitory processor-readable storage medium of claim 12 , wherein performing the one or more automated actions comprises automatically updating a device component replacement policy.

14. The non-transitory processor-readable storage medium of claim 12 , wherein the program code when executed by at least one processing device further causes the at least one processing device:

to train the one or more deep learning techniques using the at least one generated prediction.

15. The non-transitory processor-readable storage medium of claim 12 , wherein the one or more deep learning techniques comprise at least one of:

at least one deep learning technique based at least in part on a hybrid hidden Markov model; and

at least one deep learning technique based at least in part on a doubly stochastic model.

16. An apparatus comprising:

at least one processing device comprising a processor coupled to a memory;

the at least one processing device being configured:

to obtain telemetry data from one or more enterprise devices;

to determine, for each of the one or more enterprise devices, values for multiple device attributes by processing the obtained telemetry data;

to generate, for each of the one or more enterprise devices, at least one prediction related to lifecycle information of at least one device component by processing the determined attribute values using one or more deep learning techniques, wherein the one or more deep learning techniques comprise at least one neural network containing at least one transition matrix between two or more hidden states of the at least one neural network, wherein each of the two or more hidden states comprises one or more probability distributions associated with one or more input layers of the at least one neural network; and

to perform one or more automated actions based at least in part on the at least one generated prediction.

17. The apparatus of claim 16 , wherein performing the one or more automated actions comprises automatically updating a device component replacement policy.

18. The apparatus of claim 16 , wherein the at least one processing device being further configured:

to train the one or more deep learning techniques using the at least one generated prediction.

19. The apparatus of claim 16 , wherein the one or more deep learning techniques comprise at least one of:

at least one deep learning technique based at least in part on a hybrid hidden Markov model; and

at least one deep learning technique based at least in part on a doubly stochastic model.

20. The apparatus of claim 16 , wherein the one or more deep learning techniques comprise at least one deep learning technique based at least in part on a hybrid hidden Markov model integrated with a doubly stochastic model.