Research Area


Watch


Date


Author-Data Scientist

SORT:

CPI Prediction Model March 2023

Ryan Peng | Lead Data Scientist | 20 Pages | 04/13/2023

The 20-page report presents results from both univariate and multivariate models. For the univariate model, despite approximating the core CPI's shape, it underestimates some values, with a March 2023 prediction of a 304.81 CPI value, translating to a 5.45% YOY increase. On the other hand, the multivariate model exhibits improved predictions, with fewer fluctuations in loss, and a closer alignment to the actual core CPI. However, it still tends to underpredict. For March 2023, the multivariate model forecasts a 304.07 CPI basis point, corresponding to a 5.20% YOY CPI change. The quantitative differences between the two models indicate the multivariate model's superior performance in capturing the core CPI trend.

Q1 2023 MLR Report

Ryan Peng, Duy Dang | Lead Data Scientist, Data Scientist Intern | 14 Pages | 04/11/2023

Loxz Digital's Q1 2023 MLR report is a concise, 14-page overview of the current state of machine learning in business. It highlights recent advancements, such as Generative AI models like ChatGPT and Notion AI, and discusses competitive trends and best practices. The report also examines transformative technologies like LLMs for particular industries and the reverse engineering component of text to image generation to enhance LLMs. Very special thanks to our Data Science Intern Duy Dang and Ryan Peng.

CPI Prediction Model Feb 2023

Ryan Peng | Lead Data Scientist | 26 Pages | 95% Accuracy | 03/25/2023

In this 20-page report, our data science lead designs an LSTM model to predict The Core CPI. With a 95% confidence interval rate it measures changes in goods and services prices, excluding food and energy, providing a stable measure of inflation. This report is published before the government data is released and is critical for policymakers and economists guiding monetary policy decisions and influencing pricing, wages, and investments.

Smart Workflows

Loxz Digital | 03/08/2023

With Loxz Digital's smart workflows, machine learning meets business efficiency. Discover how our predictive algorithmic platform helps automate and optimize workflows for maximum efficiency and productivity

Business Value to using ML

Loxz Digital | 03/08/2023

With Loxz Digital, you can unlock the power of machine learning to drive incremental value. Learn about our comprehenisve approach to model sustainabiliy and model optimization.

Hyper-Localized Model Stack at Loxz

Loxz Digital | 03/08/2023

Unlock the power of hyper-localized insights to drive business growth with ML. Learn about our model stack and model sustainability and how it can help you make more intelligent decisions.

Guide to RealTime ML

Loxz Digital | 03/08/2023

Stay ahead of the curve with Loxz Digital's guide to real-time machine learning. Discover how our realitme predictive algorithmic platform ensures informed decisions to achieve your business objectives in real-time.

Many ways to Engage with Loxz

Loxz Digital | 03/08/2023

At Loxz Digital, we believe in the power of engagement to drive business growth. Learn about the different ways you can interact with our model portfolio and begin sharing insights with your subscribers.

Sentiment Analysis Model v1.2

Buwani Manuweera | Data Scientist | 10 Pages | 95% Accuracy | 03/04/2023

In this 10-page report, we are presenting version 1.12 of our Sentiment Analysis Model, which includes several new features to improve its accuracy and usefulness. The current version expands on an already comprehensive tool for analyzing and generating text with the desired tone. It includes new features such as tone modalities, tone sliders, generative text, Loxz Recommended tone, and region-based tone intensities. The incorporation of an LLM further enhances the model as a valuable asset for anyone looking to optimize their communication strategy.

CPI Prediction Model Jan 2023

Ryan Peng | Lead Data Scientist | 26 Pages | 95% Accuracy | 02/14/2023

The January 2023 CPI report utilizes an LTSM model that is capable of analyzing large datasets, detecting patterns, and forecasting inflation rates and core CPI. The model's performance is continually evaluated and optimized to maintain its high accuracy. The current confidence level of this model is 95%. This model utilizes historical data and trends and also includes datasets from the St. Louis Fed to generate accurate predictions. Our economic reports, which we release one or two days prior to the government's monthly reports provide insights and may be particularly useful for economists who want an early read on inflation data.

Q4 2022 ML Readiness Survey Report

Ryan Peng, Duy Dang | Lead Data Scientist, Data Scientist Intern | 02/14/2023

In this 28-page Machine Learning Readiness report, Duy Anh Dang one of our Data Science Interns and our Lead Data Scientist Ryan Peng, articulate their thoughts on identifying #ModelHubs, the robust investments in #GenerativeAI, #SmartFactories, Machine Learning as a Service along with a special section identifying trends over the next 10 years by Andrew Ng. This report will be uploaded to our resources.loxz.com app over the weekend.

Font Optimization Model

Buwani Manuweera | Data Scientist | 16 Pages | 84.98% Accuracy | 01/28/2023

Choosing a specific font for a targeted email is an important task for any email campaign. Based on the targeted group, you need to consider the font style, size, and many other properties so the users feel included and comfortable reading the content. The Loxz Font Optimization Model helps you determine the best possible font properties for your emails in a particular industry and a campaign type. We apply the Random Forest algorithm for predictions in this model and it achieves the highest accuracy of 84.98% and provides three recommendations that will increase the current engagement rate.

CPI Prediction Model December 2022

Ryan Peng | Lead Data Scientist | 24 Pages | 95% Accuracy | 01/14/2023

The Consumer Price Index (CPI) is a measure of the change in the price of a basket of goods and services consumed by households. It is commonly used as a measure of inflation, as it reflects the changing purchasing power of consumers. LSTM networks can potentially be used for predicting changes in the CPI over time. In this case, the input to the LSTM network would be a sequence of past CPI values, and the output would be a predicted future CPI value. The LSTM network can then be trained using supervised learning, where the training data includes both the input and the known correct output.

CPI Prediction Model November 2022

Miu Lun (Andy) Lau | Data Scientist | 24 Pages | 95% Accuracy | 01/12/2023

The Consumer Price Index (CPI) is a measure of the change in the price of a basket of goods and services consumed by households. It is commonly used as a measure of inflation, as it reflects the changing purchasing power of consumers. LSTM networks can potentially be used for predicting changes in the CPI over time. In this case, the input to the LSTM network would be a sequence of past CPI values, and the output would be a predicted future CPI value. The LSTM network can then be trained using supervised learning, where the training data includes both the input and the known correct output.

Q3 2022 ML Readiness Survey Report

Miu Lun (Andy) Lau | Data Scientist | 11/16/2022

Our Q3 2022 MLR Report dives into popular Frameworks, IDEs, transformers including Generative AI. We are seeing investors gravitate toward Generative AI, as the next frontier in ML.

Survey Incentive Model

Chen Song | Data Scientist | 15 Pages | 98.89% Accuracy | 9/10/2022

The Survey Incentive Model provides RealTimeML predictive analytics on the optimized survey incentives for a given email campaign. These incentives are built directly within the workflow of an email campaign and provides three recommendations for optimal engagement. Besides the survey incentive amount, it also optimizes for survey length based on type of industry and type of email campaign you are optimizing for.

Call to Action Model

Buwani Manuweera | Data Scientist | 17 Pages | 80.492% Accuracy | 9/10/2022

The Call to Action Model provides real-time predictive analytics for specific call-to-action buttons included in the email campaign prior to deploying it, in order to improve the engagement rate based color and text of the CTA. These predictions are served to the campaign engineer in milliseconds upon run, to optimize the CTAs in the email campaign prior to deployment.

Q2 2022 ML Readiness Survey Report

Chen Song | Data Scientist | 8/8/2022

Machine Learning (ML), as an optimization process for AI technologies, is vital for providing and achieving proven more efficient, and smarter AI solutions. Embracing machine learning is not optional but required nowadays because adopting machine learning drives a dramatic delta in business results that potentially improves an organization’s bottom line.

Q1 2022 MLR Report

Chen Song | Data Scientist | 4/16/2022

The 2021 Annual ML Survey Report examines the state of Machine Learning Readiness among organizations, industries and role including an introduction to our new sub-scoring methodology.

2021 Annual MLR Report

Chen Song | Data Scientist | 1/24/2022

The 2021 Annual ML Survey Report examines the state of Machine Learning Readiness among organizations, industries and role including an introduction to our new sub-scoring methodology.

Data Augmentation Model

Casey Yoon | Data Scientist | 17 Pages | 79% Accuracy | 01/21/2022

This RealTimeML Data Augmentation model provides real time predictive analytics on the images used for digital marketing campaigns. The model is essentially an image classification model that will identify the best image augmentations and alternate images and recommend these options for higher target variable conversion rates.

Q3 2021 MLR Report

Justin Chase, Chen Song | Data Scientist | 11/5/2021

This third quarter report provides a comprehensive overview of the Loxz Digital Machine Learning Readiness Survey, methodological considerations around our innovative sub-scores, and essential insights - including industry trends across the machine learning lifecycle from our latest wave of data collection.

Mining MLR Insight With Subscores

Justin Chase | Data Scientist | 10/18/2021

Never underestimate taking a bird’s eye view. In this 15-page report, we introduce the use of sub-scores in the Machine Learning Lifecycle which can provide the most comprehensive overview when determining your company’s ML readiness. The report is designed to be a pre-cursor to our Q3, 2021 Report.

MLR Loxz Scoring Methodology

Chen Song | Data Scientist | 8/27/2021

It’s a diagnostic assessment which uses a tightly vetted answer key system to simultaneously increase accuracywhile reducing bias. It is meant to assess the readiness of an organization that is considering building ML Models.

MLR Pearson Methodology

Yiming Zhang | Data Scientist | 4/13/2022

In this methodology paper, we will dive into how to generate rich insights from a survey respondent to compare their specific score with a unique user group that represents the respondent's profile.

MLR: Semantic Scoring Methodology

Yiming Zhang | Data Scientist | 16 Pages | 5/18/2022

In this methodology paper, we will dive into how to generate rich insights from a survey respondent to compare their specific score with a unique user group that represents the respondent's profile.

Risk Perspective

Chen Song | Data Scientist | 7/13/2021

In this 2-page report discover whether your company segment takes enough risk when building models. Leaders tend to be more conservative while start ups take more risk.

RealTimeML Character Count Model

Chen Song | Data Scientist | 9 Pages | 85% Accuracy | 5/23/2022

As there are no existing datasets directly related to email content available publicly across industries and target variables (open rate, click-through rate, unsubscribed rate and bounce rate, etc), we collected different email campaigns from 9 different industries, including academic and education, energy, entertainment, finance and banking, healthcare, hospitality, real estate, retail, and software and technology

Sentiment Analysis Model

Jeffery Ott | Data Scientist | 12 Pages | 95% Accuracy | 8/8/2022

The model is a two-part model, meaning during the first step, the text from email is parsed and loaded into our sentiment model. Part one is a BERT model, which analyzes the text in a bidirectional format and returns a multi-label classification of 8 different tones. These tones will evolve over time and subsequent builds of the model can be tested with different tones based on client requirements.

ShotSpotter Model

Miu Lun (Andy) Lau, Chen Song | Data Scientist | 15 Pages | 72.2% Accuracy | 07/21/2022

This model is the first model designed for Law Enforcement. It seeks to serve predicts future potential locations for gunfire. Using data from ShotSpotter API, this spatio-temporal model uses data from audio sensing equipment and immediately retrains the model to serve new locations where future gunshots will be fired.

Discount Optimization Model For Email

Buwani Manuweera | Data Scientist | 17 Pages | 92.8% Accuracy | 07/11/2022

The Discount Optimization model is developed to provide predictive analytics on product discounts offered in email campaigns. The model provides predictions on an email campaign for the selected esmail engagement metrics as well as recommendations on how to optimize the discounts in emails in order to maximize the specified targeted campaign metrics.

Image Optimization Model

Miu Lun (Andy) Lau | Data Scientist | 15 Pages | 91% Accuracy | 06/10/2022

The goal of this ML model is to provide RealTime Predictive Analytics on the recommended images to provide accurate email engagement metrics and timely feedback for email related marketing campaigns. We utilizes an automotive image dataset and generate additional images using data augmentation technique and pipeline. This particular prediction model utilizes the CNN image based ResNet model, and we performed transfer learning using the existing parameters and weights from an ImageNet dataset.

Q2 2021 MLR Survey Report

Chen Song | Data Scientist | 7/7/2021

This report released on July 7, 2021, focuses on whether a company is ready for the implementation of machine learning in their organization and is structured to help you define four major barriers to assist in your efforts.