International Business Machines Corporation

Petr Novotny

Nitin Gaur

Qi Zhang

Lei Yu

Timothy Olson

Dulce Beatriz Ponceleon

Salman Abdul Baset

Yacov Manevich

Jeronimo Irazabal

Venkatraman Ramakrishna

Hong Min

Donna N Dillenberger

Palani Sakthi

Artem Barger

Shaikh Shahriar Quader

Sarbajit Kumar Rakshit

Shikhar Kwatra

Ashish Kundu

Dushyant K Behl

Chander Govindarajan

Matthias Seul

Bishakh Chandra Ghosh

Aindrila Basak

Daniel W Bingham

Ravi Nair

Raghuveer Prasad Nagar

Timothy S Olson

Shyam Ramji

Motohiro Kawahito

Takuya Nakaike

Yada Zhu

Berthold Reinwald

Yuan Yuan

Mathew S Thoennes

Taesung Lee

Horst Cornelius Samulowitz

Amitava Kundu

Shiqiang Wang

Donna N Eng Dillenberger

Luis Angel Bathen

Sean K Dunne

Mariela Claudia Lanza

Wisam Mohammed

Adrian Mahjour

Chad Marston

Dheeraj Arremsetty

Carlo Appugliese

In accordance with an embodiment of the invention, a method is provided for performing data pre-processing operations during data preparation of a machine learning lifecycle. The method includes defining one or more data pre-processing functions for applying to data stored in a dataset, executing one or more learn functions for learning the data, and executing one or more transform functions for transforming the data. Each of the one or more learn functions generates a first Structured Query Language (SQL) statement representing a definition of corresponding learn function for corresponding defined data pre-processing function. Each of the one or more transform functions generates a second SQL statement representing a definition of corresponding transform function for corresponding defined data pre-processing function. The dataset is stored in a database.

Performing data pre-processing operations during data preparation of a machine learning lifecycle

Providing a representative dataset from an initial dataset by accessing a dataset associated with a machine learning model, receiving input parameters associated with the representative dataset selection, the input parameters including an evaluation metric, determining a density of a plurality of datapoints associated with the dataset, training a first iteration of a machine learning model using a first data point selected according to the density, determining a first value of the evaluation metric for the first iteration of the machine learning model, generating a representative subset based on the first value of the evaluation metric value, and providing the representative dataset and a final machine learning model trained using the representative dataset.

Selecting a high coverage dataset

An example operation may include one or more of generating a blockchain request which comprises data for a blockchain, creating a tag value which identifies whether the blockchain request is a noisy request that comprises fake data or a non-noisy request that comprises non-noisy data, storing the tag value within the blockchain request, and transmitting the blockchain request to one or more blockchain peers.

Noisy transaction for protection of data

A system may enact a process including initiating an operation to transfer a secret, establishing n groups of nodes in a blockchain network, dividing a secret into n secret parts, creating n group keys, encrypting each secret part of the secret with a corresponding group key, and recording encrypted secret parts on the blockchain networks.

Secret smart operations in blockchain

A display case is provided. The display case includes a transparent display for viewing an object through the transparent display and a watermarking engine configured to display a watermark on the transparent display.

Display case with transparent display for watermarking

An example operation may include one or more of transmitting a user identifier to a plurality of blockchain peers of a blockchain network, receiving a plurality of public keys from the plurality of blockchain peers, receiving a plurality of ciphertexts from the plurality of blockchain peers generated based on the user identifier and a corresponding plurality of private keys, generating a combined ciphertext that includes the plurality of ciphertexts and generating a temporary blockchain address of the blockchain network based on the plurality of public keys, and storing the combined ciphertext and a transfer of an asset via the temporary blockchain address of the blockchain network.

Future asset reclamation via blockchain

An example operation may include one or more of receiving, via a first blockchain peer, a request to verify an identity of a user, performing, via the first blockchain peer, an identity validation of the user via a know your customer (KYC) service provider of a blockchain network, determining whether a digital identity of the user has been previously stored by another blockchain peer on a blockchain of the blockchain network, in response to a determination that the digital identity of the user has not been stored on the blockchain, creating a non-fungible token (NFT) that identifies the user, and appending information about the identity validation of the user to a structure of the NFT and storing the NFT in a block of the blockchain.

Token-based identity validation via blockchain

A node in a blockchain network may agree, on an authority accept a compliance module from the authority, accept the compliance module. The node may also receive an operation, verify a compliance of the operation based on the compliance module, add the verified operation to a ledger on the blockchain network.

Compliance mechanisms in blockchain networks

An example operation includes one or more of detecting, by a validation service node, an access of a block on a blockchain ledger; accessing, by the validation service node, a block validation stamp comprising meta-data signed by a trusted stamping authority (TSA) on the blockchain ledger; and validating the block based on the meta-data.

Trusted ledger stamping

An example operation may include one or more of encrypting a private key with an encryption key, generating a plurality of keys based on the encryption key and converting the plurality of keys into a plurality of key shares based on a secret input value, storing the encrypted private key on a blockchain, and distributing the plurality of key shares to a plurality of blockchain peers of the blockchain, where the distributing comprises transmitting a different key share from among the plurality of key shares to each blockchain peer among the plurality of blockchain peers.

Key reclamation in blockchain network via OPRF

An example operation may include one or more of receiving a request for storage at a blockchain network, attaching, via a blockchain node, a verifiable credential created by a self-sovereign identity (SSI) network to a blockchain transaction associated with the request, where the verifiable credential includes a claim of the blockchain node and a proof of the SSI network that created the verifiable credential, transmitting the blockchain transaction and the attached verifiable credential to one or more other blockchain nodes, and storing the blockchain transaction and the attached verifiable credential via a data block on the blockchain.

Blockchain network identity management using SSI

Mechanisms for performing intelligent federated machine learning (ML) model updates are provided. A plurality of ML model updates, and a plurality of dataset sketch commitment data structures (sketches), are received from a plurality of participant computing systems. Each sketch provides statistical characteristics of a corresponding local dataset used by a corresponding participant to train a local ML model. A potentially malicious participant identification operation is performed based on an analysis of the plurality of sketches to identify one or more potentially malicious participants based on differences in sketches. ML model updates received from participant computing systems identified as potentially malicious participants are discarded to thereby generate a modified set of updates. The federated ML computer model is updated based on the modified set of updates.

Federated learning with dataset sketch commitment based malicious participant identification

Embodiments of the present invention provide methods, computer program products, and systems. Embodiments of the present invention can receive, by a computing device, a request to access a datapoint of a machine learning dataset contained in a database. Embodiments of the present invention can access, by the computing device, a virtual data frame that includes a schema which represents a structure of the machine learning dataset in the database. Embodiments of the present invention can retrieve, by the computing device, the datapoint of the machine learning utilizing the virtual data frame and return, by the computing device, the retrieved datapoint in response to the request.

Processing large machine learning datasets

An example operation may include one or more of receiving, via a custodial service, a request to transact with a digital asset owned by a user and temporarily in custody of the custodial service, generating a blockchain transaction comprising an identifier of the digital asset on a blockchain ledger, an identifier of the custodial service, and an identifier of a recipient of the digital asset, signing the blockchain transaction with a key from a trifocal key which proves that the custodial service is authorized to transact with the digital asset on behalf of the user, and storing the signed blockchain transaction on a blockchain ledger.

Trifocal key for controlling custodians of digital assets

A node in a blockchain network may receive an event initiation for a first digital corollary of a first physical asset, receive conditional data related to the first digital corollary, detecting, by the node, a link to a second digital corollary, prompt a simulation of the second digital corollary, receive an outcome of the simulation, and determine an executable action for the second digital corollary.

Glue-contracts for digital corollaries in blockchain networks

Growing community of inventors

Mount Kisco, NY, United States of America

Petr Novotny