Methods and apparatuses for AI-based ledger prediction
Apparatuses and methods for AI-based ledger prediction are provided. A processor is configured by instructions on a memory to receive and categorize a ledger file to a ledger type based on the data contained within the ledger file. The ledger data may contain information related to an insurance policy and investments made based on the payments into the policy over a period of time. The processor may be configured to utilize machine learning to generate a prediction of a value or values related to the ledger data, for example a retirement distribution amount.
1. An apparatus for predictive ledger generation, the apparatus comprising:
a processor; and
a memory communicatively coupled with the processor, the memory containing instructions stored thereon, the instructions configuring the processor to:
receive a ledger file containing ledger data;
classify the ledger file to a ledger type, wherein classifying the ledger file further comprises:
receiving ledger training data correlating a plurality of ledger data types with a plurality of ledger classification types;
training a ledger classification machine learning model with the ledger training data;
outputting the one or more ledger classifications by:
inputting the ledger file into the trained ledger classification machine learning model; and
receiving the one or more segment trendlines as outputs from the trained ledger classification machine learning model; and
identify one or more trends in the ledger data, wherein identifying the one or more trends in the ledger data further comprises:
segmenting the ledger data into one or more trend segments based on the identified one or more trends
fitting one or more segment trendlines to each of the one or more trend segments, wherein fitting the one or more segment trendlines to each of the one or more trend segments further comprises:
receiving trendline training data correlating a plurality of trendline types with a plurality of segment types;
training a trendline fit machine learning model with the trendline training data; and
outputting the one or more segment trendlines by:
inputting the trend segments to the trained trendline fit machine learning model; and
receiving the one or more segment trendlines as outputs; and
generating at least one predictive trendline based on the segment trendlines.
2. The apparatus of claim 1 , wherein the processor is configured to optimize the at least one predictive trendline based on input received from a user through a user interface.
3. The apparatus of claim 1 , wherein the processor is further configured to:
identify one or more external events influencing the ledger data; and
segment the ledger data into one or more trend segments as a function of the one or more external events.
4. The apparatus of claim 3 , wherein the processor is further configured to minimize an effect of the one or more external events on the ledger data.
5. The apparatus of claim 1 , wherein the processor is further configured to select one or more elements of the ledger data to store to the blockchain based on the ledger type.
6. The apparatus of claim 1 , wherein the processor is further configured to classify the ledger file to a ledger type based on a pecuniary parameter.
7. The apparatus of claim 6 , wherein the pecuniary parameter corresponds to both of an insurance parameter and an investment parameter.
8. A method for predictive ledger generation, the method comprising:
receiving, by a processor, a ledger file containing ledger data;
classifying, by the processor, the ledger file to a ledger type, wherein classifying the ledger file to the ledger type further comprises:
receiving, by the processor, ledger training data correlating a plurality of ledger data types with a plurality of ledger classification types;
training, by the processor, a ledger classification machine learning model with the ledger training data; and
outputting, by the processor, the one or more ledger classifications by:
inputting the ledger file into the trained ledger classification machine learning model; and
receiving the one or more segment trendlines as outputs from the trained ledger classification machine learning model; and
analyzing, by the processor, the ledger file to identify one or more trends in the ledger data, wherein analyzing the ledger file further comprises:
segmenting, by the processor, the ledger data into one or more trend segments based on the identified one or more trends;
fitting, by the processor, one or more segment trendlines to each of the one or more trend segments, wherein fitting the one or more segment trendlines further comprises:
receiving, by the processor, trendline training data correlating a plurality of trendline types with a plurality of segment types;
training, by the processor, a trendline fit machine learning model with the trendline training data; and
outputting, by the processor, the one or more segment trendlines by:
inputting the trend segments to the trained trendline fit machine learning mode; and
receiving the one or more segment trendlines as outputs from the trained trendline fit machine learning model; and
generating, by the processor, at least one predictive trendline based on the segment trendlines.
9. The method of claim 8 , further comprising optimizing, by the processor, the at least one predictive trendline based on input received from a user through a user interface.
10. The method of claim 8 , further comprising segmenting, by the processor, the ledger data into one or more trend segments by identifying one or more external events influencing the ledger data.
11. The method of claim 10 , further comprising minimizing, by the processor, an effect of the one or more external events on the ledger data.
12. The method of claim 8 , further comprising selecting, by the processor, one or more elements of the ledger data to store to the immutable sequential listing based on the ledger type.
13. The method of claim 8 , further comprising classifying, by the processor, the ledger file to a ledger type based on a pecuniary parameter.
14. The method of claim 13 , wherein the pecuniary parameter corresponds to both of an insurance parameter and an investment parameter.
15. The apparatus of claim 1 , wherein fitting the one or more segment trendlines to each of the one or more trend segments further comprises:
determining a plurality of coefficients for the one or more fitted segment trendlines by iterating an optimization;
generating a plurality of weighted segment trendlines with the plurality of coefficients;
training the trendline fit machine learning model iteratively with the trendline training data and the plurality of weighted segment trendlines.
16. The method of claim 8 , wherein fitting the one or more segment trendlines further comprises:
determining a plurality of coefficients for the one or more fitted segment trendlines by iterating an optimization;
generating a plurality of weighted segment trendlines with the plurality of coefficients;
training the trendline fit machine learning model iteratively with the trendline training data and the plurality of weighted segment trendlines.
17. The apparatus of claim 1 , wherein the ledger file is stored to an immutable sequential listing as a function of the ledger type and the identified one or more trends.
18. The method of claim 8 , wherein the ledger file is stored, by the processor, to an immutable sequential listing as a function of the ledger type and the identified one or more trends.