Sap Se

Philipp Knuesel

Roman Rommel

Janick Frasch

Santo Bianchino

The present disclosure involves systems, software, and computer implemented methods for evaluating machine learning on remote datasets using confidentiality-preserving evaluation data. In response to determining that data of the remote customer dataset is of sufficient quality and quantity, feature data corresponding to a machine learning pipeline is generated. The remote customer dataset into one or more data partitions and for each partition, one or more baseline models and one or more machine learning models are trained using a machine learning library included in the remote customer database. Aggregate evaluation data is generated for each baseline model and each machine learning model that includes model debrief data and customer data statistics. In response to determining that the customer has enabled sharing of the aggregate evaluation data with a software provider who provided the remote customer database, the aggregate evaluation data is provided to the software provider.

Evaluating machine learning on remote datasets using confidentiality-preserving evaluation data

Computer-readable media, methods, and systems are disclosed for scheduling a start time and a shutdown time of one or more online resources associated with a multi-cloud resource scheduler. A request from a first user is received to access a multi-cloud resource scheduler associated with one or more online resources. Responsive to the request from the first user, credentials of the first user are validated prior to providing access to the multi-cloud resource scheduler. Based upon validating the credentials of the first user, access to the multi-cloud resource scheduler is provided. Instructions are received from the first user to schedule a start time and a shutdown time of at least one online cloud resource connected to the multi-cloud resource scheduler. An availability of the at least one online cloud resource is established for access by a second user based on the instructions.

Multi-cloud resource scheduler

The present disclosure involves systems, software, and computer implemented methods for data confidentiality-preserving machine learning on remote datasets. An example method includes receiving connection information for connecting to a remote customer database and storing the connection information in a machine learning runtime. Workload schedule information for allowable time windows for machine learning pipeline execution on remote customer data of the customer is received from the customer. A determination is made that an execution queue includes a machine learning pipeline during an allowed time window. The connection information is used to connect to the remote customer database during the allowed time window. Execution is triggered by the machine learning runtime of the machine learning pipeline on the remote customer database. Aggregate evaluation data corresponding to the execution of the machine learning pipeline on the remote customer database is received and provided to a user.

Growing community of inventors

Nussloch, Germany

Philipp Knuesel