Cold start machine learning algorithm

1. A computerized method of selecting one or more images from a plurality of images, the method comprising:

identifying a first plurality of images stored on a computing device, each of the first plurality of images having an indication that they depict a first member of a social networking service;

using the first plurality of images as feature data as input to train a first machine learning algorithm model that is unique to the first member, the first machine learning algorithm model designed to calculate a member likelihood score for a candidate image based on the feature data, the member likelihood score being a measurement of a likelihood that the candidate image depicts the first member;

obtaining a second plurality of images stored on the computing device, at least some of the second plurality of images not having the indication that they depict the first member;

feeding each of the second plurality of images to the first machine learning algorithm model corresponding to the first member, obtaining a member likelihood score for each of the second plurality of images; and

based on the member likelihood scores for the second plurality of images, selecting one or more member images from the second plurality of images.

2. The computerized method of claim 1 , further comprising:

passing the selected one or more member images to a second machine learning algorithm model, the second machine learning algorithm model trained to output a professionalism score for each image passed to it; and

using the professionalism score of at least one of the selected one or more member images to perform a transformation of the at least one of the selected one or more member images to improve the professionalism score.

3. The method of claim 1 , wherein the indication that the first plurality of images depicts the first member is that they are stored in a folder labeled in a manner suggesting that they depict the first member.

4. The method of claim 1 , wherein the first machine learning algorithm model is a deep convolutional neural network.

5. The method of claim 2 , wherein the second machine learning algorithm model is a deep convolutional neural network.

6. The method of claim 1 , wherein the computing device is a mobile device having a front-facing camera.

7. The method of claim 1 , wherein the selecting includes selecting any of the second plurality of images having member likelihood scores transgressing a preset threshold.

8. A system comprising:

one or more processors;

a non-transitory computer-readable medium having instructions stored thereon, which, when executed by the one or more processors a processor, cause the system to:

identify a first plurality of images stored on a computing device, each of the first plurality of images having an indication that they depict a first member of a social networking service;

use the first plurality of images as training data to a first machine learning algorithm to train a first machine learning algorithm model uniquely corresponding to the first member, the first machine learning algorithm model corresponding to the first member and designed to calculate a member likelihood score for a candidate image, the member likelihood score being a measurement of a likelihood that the candidate image depicts the first member;

obtain a second plurality of images stored on the computing device, at least some of the second plurality of images not having the indication that they depict the first member;

feed each of the second plurality of images to the first machine learning algorithm model corresponding to the first member, obtaining a member likelihood score for each of the second plurality of images; and

based on the member likelihood scores for the second plurality of images, select one or more member images from the second plurality of images.

9. The system of claim 8 , wherein the instructions further cause the system to:

pass the selected one or more member images to a second machine learning algorithm model, the second machine learning algorithm model trained to output a professionalism score for each image passed to it; and

use the professionalism score of at least one of the selected one or more member images to perform a transformation of the at least one of the selected one or more member images to improve the professionalism score.

10. The system of claim 8 , wherein the indication that the first plurality of images depicts the first member is that they are stored in a folder labeled in a manner suggesting that they depict the first member.

11. The system of claim 8 , wherein the first machine learning algorithm model is a deep convolutional neural network.

12. The system of claim 9 , wherein the second machine learning algorithm model is a deep convolutional neural network.

13. The system of claim 8 , wherein the computing device is a mobile device having a front-facing camera.

14. The system of claim 13 , wherein the selecting includes selecting any of the second plurality of images having member likelihood scores transgressing a preset threshold.

15. A non-transitory machine-readable storage medium comprising instructions, which, when implemented by one or more machines, cause the one or more machines to perform operations comprising:

identifying a first plurality of images stored on a computing device, each of the first plurality of images having an indication that they depict a first member of a social networking service;

using the first plurality of images as training data to a first machine learning algorithm to train a first machine learning algorithm model uniquely corresponding to the first member, the first machine learning algorithm model corresponding to the first member and designed to calculate a member likelihood score for a candidate image, the member likelihood score being a measurement of a likelihood that the candidate image depicts the first member;

obtaining a second plurality of images stored on the computing device, at least some of the second plurality of images not having the indication that they depict the first member;

feeding each of the second plurality of images to the first machine learning algorithm model corresponding to the first member, obtaining a member likelihood score for each of the second plurality of images; and

based on the member likelihood scores for the second plurality of images, selecting one or more member images from the second plurality of images.

16. The non-transitory machine-readable storage medium of claim 15 , further comprising:

passing the selected one or more member images to a second machine learning algorithm model, the second machine learning algorithm model trained to output a professionalism score for each image passed to it; and

using the professionalism score of at least one of the selected one or more member images to perform a transformation of the at least one of the selected one or more member images to improve the professionalism score.

17. The non-transitory machine-readable storage medium of claim 15 , wherein the indication that the first plurality of images depict the first member is that they are gored in a folder labeled in a manner suggesting that they depict the first member.

18. The non-transitory machine-readable storage medium of claim 15 , wherein the first machine learning algorithm model is a deep convolutional neural network.

19. The non-transitory machine-readable storage medium of claim 16 , wherein the second machine learning algorithm model is a deep convolutional neural network.

20. The non-transitory machine-readable storage medium of claim 15 , wherein the computing device is a mobile device having a front-facing camera.