Caldon, Giuliano

A machine learning based (ML-based) method and system for forecasting cash flow, is disclosed. Initially, the data associated with business units are obtained from data sources. The data are pre-processed to generate pre-processed data. Features associated with financial information are determined for horizons based on the pre-processed data using a plurality of input AI models. Feature combinations are generated by integrating the features associated with the financial information, for each horizon of the horizons. Forecasts are generated for each horizon for a pre-determined time interval, using a stacked AI model including forecasting models. The generated forecasts for the cash flow of the business units, are provided as an output, to the users on user interfaces associated with electronic devices associated with the users.

A machine-learning based (ML-based) system and method for automatically extracting one or more data fields from one or more documents, are disclosed. The ML-based system includes a document obtaining subsystem to obtain documents, a document pre-processing subsystem to generate pre-processed data, a field identifying subsystem to identify data fields using a trained ML model, and a field extracting subsystem to extract financial information. The ML-based system also comprises an output subsystem to deliver the extracted data to end users via user interfaces. The ML model is trained using historical documents, labelled data fields, and features such as distance-based features, direction-based features, dimension-based features, positional features, and value-based features. The M-based system employs hyperparameter optimization, noise removal, and accuracy assessment mechanisms to enhance performance. This ML-based system provides a scalable, accurate, and automated solution for financial information extraction, ensuring efficiency, adaptability, and seamless integration with enterprise systems.

A machine learning based (ML-based) method and system for redistributing data, is disclosed. Initially, an input data associated actual bank cash is obtained from data sources. The input data is pre-processed to generate pre-processed data. A month level data associated with the actual bank cash is predicted for a pre-determined horizon based on at least one of: historical cash flow data and seasonality, using machine learning (ML) models on the pre-processed data. At least one of: the month level data to week of month (WOM) level data and the WOM level data to day level data, is redistributed based on a pro-rata configuration using hyperparameters. At least one of: the WOM level data and the day level data, redistributed from the month level data, is provided as an output, to the users on user interfaces associated with electronic devices associated with the users.

Hydraulic binder composition, which comprises steel slag or a mixture consisting of steel slag and calcium carbonate, and at least one modifier additive, modifying the setting reaction of said steel slag, in a quantity between 0.1 and 20% by weight with respect to the weight of the steel slag.

The present invention discloses a button edge-cutting can opener, comprising a main body (100), a button (200), a driving unit (300), and a cutting unit (400), wherein the button (200) is rotatably connected to the main body (100) through the driving unit (300) and provided with a pressing portion (210), the cutting unit (400) is mounted on the main body (100), the driving unit (300) comprises a cone rod (310), which has a driving shaft (311) and a cone disc (312), the cone disc (312) is connected to one end of the driving shaft (311), a boss (313) is arranged on the cone disc (312), 28 to 36 gear teeth (314) are distributed on a side wall of the boss (313) in an evenly spaced manner, an end portion of each of the gear teeth (314) extends towards the cone disc (312), and a smooth surface (315) is formed between the extended end portion and an edge of the cone disc (312). The button edge-cutting can opener in the present invention locks or unlocks a canned product by pushing the pressing portion (210) upward or pressing it downward with a thumb. With the arc-shaped end portions of identification positioning backup plates (510), it is easy for a first-time user to learn and master, and the operation is convenient and reliable. Moreover, by reducing the number of the gear teeth to 28 to 36, the teeth marks generated by front and rear gear teeth can overlap, thereby reducing displacement and sliding. As such, iron filings that fall from the can edge of the scratched can lid are effectively reduced, and the food safety is ensured.

A machine learning based (ML-based) computing method and system for automatically identifying equivalent entities, is disclosed. The method involves obtaining transaction data, standardizing data fields, and generating unique transaction numbers. Entities with identical transaction numbers are grouped into initial groups. An ML model, utilizing an apriori algorithm, eliminates certain groups to form intermediate groups based on frequent item sets. An iterative unionization process then merges intermediate groups to form resultant groups of equivalent entities. The method further includes assessing model accuracy, with optional re-training if accuracy falls below a threshold. Standardization steps include column name alignment, datetime and numeric conversion, currency conversion, intra-company removal, and write-off removal. Resultant groups are provided to end users via interfaces, allowing accurate, automated identification of equivalent entities in financial transactions.

Chair (1) for a user's sitting, which comprises a support structure (2), intended to be abutted against the ground and provided with a seat (21), a back (3), mounted on the support structure (2), and provided with a first half-shell (31), which is provided with a peripheral edge (310) and has a first face (311) and an opposite second face (312), and a second half-shell (32), which is mechanically mounted on the second face (312) of said first half-shell (31), defining, with said first half-shell (31), a housing volume (V). The chair (1) also comprises a covering element (4), which comprises a upholstery (40) mounted to enclose at least the first face (311) of the first half-shell (31), and a retention band (42), fixed along the perimeter edge (400) of said covering (4). The first half-shell (31) has, on its second face (312), a plurality of first retention elements (5), and a plurality of second retention elements (6), which are interposed between the first retention elements (5) and the peripheral edge (310) of the first half-shell (31). The second retention elements (6) delimit, with said first retention elements (5), a housing channel (7), in which the retention band (42) of said covering element (4) is inserted, in order to constrain, at least temporarily, the retention element (4) to the first half-shell of the back (3) before or during the connection between the first half-shell (31) and the second half-shell (32).

Sowing device for a seeding machine, adapted to distribute seeds on the ground, which comprises a support body (2), intended to be fixed to a seeding machine and provided with at least one deposit opening (20), a hopper (3), mounted on the support body (2), adapted to contain seeds to be distributed, and provided with a pick-up opening (31). The device (1) also comprises a distribution group (4), which is mounted on the support body (2), and is configured for receiving at least one seed from the hopper (3), through the pick-up opening (31), and moving it to the deposit opening (20) of the support body (2). The sowing device (1) also comprises at least one vibration generator (5), mechanically connected to the hopper (3) and adapted to impart a vibration to the hopper (3), in order to induce the decompaction of said seeds in proximity to the pick-up opening (31).

Selector machine comprising conveyance means (2) which define an advancement path (A) along which a loose product advances, comprising multiple solid elements to be selected. The advancement path (A) comprises a tilted section (A1) and a subsequent analysis section (A2), wherein the loose product it is susceptible of advancing via falling. Such conveyance means (2) comprise a slide (3) placed along at least part of the tilted section (A1) and provided with a front face (4), along which the loose product is susceptible of sliding subject to the force of gravity, and an opposite rear face (5). The slide (3) at least partially delimits a front zone (Z1), on which the front face (4) faces, and a rear zone (Z2), on which the rear face (5) faces. In addition, an optical detection system (8) is provided, comprising emitter means (9), arranged for emitting electromagnetic radiations towards the analysis section (A2), and at least one optical sensor (10), arranged for capturing electromagnetic radiations coming from the loose product irradiated by the emitter means (9) and for generating measurement signals. The machine (1) additionally comprises an electronic control unit (11), arranged for analyzing the measurement signals and for sending command signals to expulsion means (12) arranged for eliminating, from the advancement path (A), determined solid elements of the loose product. The selector machine (1) also comprises a suction device (6) having a suction mouth (7) thereof placed in the rear zone (Z2) and the conveyance means (2) are provided with a first passage opening (13), which is placed at the tilted section (A1) and places the suction mouth (7) in fluid communication with the front zone (Z1), in order to allow the suction mouth (7) to suction, through the first passage opening (13), the fine dust from the loose product.

Device for modulating a laser beam, comprising a deformable mirror (2) provided with a reflective surface (21) adapted to receive an incident laser beam (101) and to reflect a corresponding reflected laser beam (102) provided with an output wavefront (FU) and actuator means operatively associated with said reflective surface (21) and actuatable for deforming said reflective surface (21). The modulation device (1) also comprises detection means (3) provided with an electromagnetic radiation source (31) adapted to emit a control beam (311) incident on the reflective surface (21), and a wavefront sensor (32) placed to intercept the reflected control beam (312) reflected by the reflective surface (21) in order to detect measurements of the wavefront of the reflected control beam (312) itself and generating corresponding measurement signals (Sm). In addition, the modulation device comprises a logic control unit (4), which receives the measurement signals (Sm), is arranged for associating the latter with corresponding forms of the output wavefront (FU) of the reflected laser beam (102), is operatively connected to the actuator means in order to actuate them to deform the reflective surface (21) in order to feedback control the form of the output wavefront (FU) of the reflected laser beam (102).