This paper offers a thorough examination of the univariate predictability in cryptocurrency time-series. By exploiting a combination of complexity measure and model predictions we explore the cryptocurrencies time-series forecasting task focusing on the exchange rate in USD of Litecoin, Binance Coin, Bitcoin, Ethereum, and XRP. On one hand, to assess the complexity and the randomness of these time-series, a comparative analysis has been performed using Brownian and colored noises as a benchmark. The results obtained from the Complexity-Entropy causality plane and power density spectrum analysis reveal that cryptocurrency time-series exhibit characteristics closely resembling those of Brownian noise when analyzed in a univariate context. On the other hand, the application of a wide range of statistical, machine and deep learning models for time-series forecasting demonstrates the low predictability of cryptocurrencies. Notably, our analysis reveals that simpler models such as Naive models consistently outperform the more complex machine and deep learning ones in terms of forecasting accuracy across different forecast horizons and time windows. The combined study of complexity and forecasting accuracies highlights the difficulty of predicting the cryptocurrency market. These findings provide valuable insights into the inherent characteristics of the cryptocurrency data and highlight the need to reassess the challenges associated with predicting cryptocurrency's price movements.

Large Language Models (LLMs) excel in diverse applications including generation of code snippets, but often struggle with generating code for complex Machine Learning (ML) tasks. Although existing LLM single-agent based systems give varying performance depending on the task complexity, they purely rely on larger and expensive models such as GPT-4. Our investigation reveals that no-cost and low-cost models such as Gemini-Pro, Mixtral and CodeLlama perform far worse than GPT-4 in a single-agent setting. With the motivation of developing a cost-efficient LLM based solution for solving ML tasks, we propose an LLM Multi-Agent based system which leverages combination of experts using profiling, efficient retrieval of past observations, LLM cascades, and ask-the-expert calls. Through empirical analysis on ML engineering tasks in the MLAgentBench benchmark, we demonstrate the effectiveness of our system, using no-cost models, namely Gemini as the base LLM, paired with GPT-4 in cascade and expert to serve occasional ask-the-expert calls for planning. With 94.2\% reduction in the cost (from \$0.931 per run cost averaged over all tasks for GPT-4 single agent system to \$0.054), our system is able to yield better average success rate of 32.95\% as compared to GPT-4 single-agent system yielding 22.72\% success rate averaged over all the tasks of MLAgentBench.

This paper introduces a novel approach to enhancing closed-domain Question Answering (QA) systems, focusing on the specific needs of the Lawrence Berkeley National Laboratory (LBL) Science Information Technology (ScienceIT) domain. Utilizing a rich dataset derived from the ScienceIT documentation, our study embarks on a detailed comparison of two fine-tuned large language models and five retrieval-augmented generation (RAG) models. Through data processing techniques, we transform the documentation into structured context-question-answer triples, leveraging the latest Large Language Models (AWS Bedrock, GCP PaLM2, Meta LLaMA2, OpenAI GPT-4, Google Gemini-Pro) for data-driven insights. Additionally, we introduce the Aggregated Knowledge Model (AKM), which synthesizes responses from the seven models mentioned above using K-means clustering to select the most representative answers. The evaluation of these models across multiple metrics offers a comprehensive look into their effectiveness and suitability for the LBL ScienceIT environment. The results demonstrate the potential benefits of integrating fine-tuning and retrieval-augmented strategies, highlighting significant performance improvements achieved with the AKM. The insights gained from this study can be applied to develop specialized QA systems tailored to specific domains.

This study explores the impact of upselling on user engagement. We model users' deposit behaviour on the fantasy sports platform Dream11. Subsequently, we develop an experimental framework to evaluate the effect of upselling using an intensity parameter. Our live experiments on user deposit behaviour reveal decreased user recall with heightened upselling intensity. Our findings indicate that increased upselling intensity improves user deposit metrics and concurrently diminishes user satisfaction and conversion rates. We conduct robust counterfactual analysis and train causal meta-learners to personalise users' upselling intensity levels to reach an optimal trade-off point.

BATON ROUGE, LOUISIANA – Three police officers were shot to death and three others wounded in Baton Rouge, Louisiana, on Sunday, in what authorities said was an ambush less than two weeks after a black man was killed by police, sparking nationwide protests. The officers in Baton Rouge were responding to a call of shots fired when they were ambushed by at least one gunman, Baton Rouge Mayor Kip Holden said. One suspect is dead and police are checking the shooting scene with a robot to make sure there are no explosives, Baton Rouge Police spokesman L'Jean Mckneely said. Police told reporters authorities are seeking more than one suspect and said the public should be on the lookout for people dressed in black and carrying long guns. President Barack Obama has been briefed on the shootings and will be updated throughout the day, the White House said.

The Web app, which is powered partly by artificial intelligence, analyzes posts on social media as well as police radio chatter and feeds of the local airspace in virtually any region. The software, which is linked to IBM's Watson artificial intelligence, combs through tweets and images, specific hashtags and phrases, or posts from or about a particular geographic area and then uses computer algorithms to gauge the mood of that swirling digital conversation. The AI aspects of the iAWACS app only monitor the social media posts -- they don't analyze the audio from police scanners nor from the airspace maps. The result, which the Jester said was still a work in progress, was built from the ground up for law enforcement and intelligence officials with real-time information needs.

As the country reels from the spasm of gun violence that killed two black men and five officers this week, a prominent digital vigilante is using an online tool he hacked together to keep an eye on hot spots that seem at risk of boiling over into bloodshed. The Web app, which is powered partly by artificial intelligence, analyzes posts on social media as well as police radio chatter and feeds of the local airspace in virtually any region. To detect rumblings of unrest and alert the public. On a recent night, the tool had its gaze trained on Baton Rouge, La., where protesters backed by the New Black Panther Party gathered for a rally. "I'm looking for any indication they are coordinating skirmishes. Using IBM's Watson AI, the tool not only examines large collections of tweets but -- somewhat eerily -- also can go through a single user's timeline and, with Watson's machine learning technology, offer an analysis of that user's "trustworthiness, propensity toward violence [and] openness," the Jester said. That information, he said, could hold clues to a criminal's intentions. If the Jester's name sounds familiar, that's because the hacker has appeared elsewhere -- on Time's list of most influential internet personalities, on CNN and, according to a recent blog post, on an upcoming episode of USA's "Mr.

As the country reels from the spasm of gun violence that killed two black men and five police officers this week, a prominent digital vigilante is using an online tool he hacked together to keep an eye on hot spots that seem at risk of boiling over into bloodshed. The Web app, which is powered partly by artificial intelligence, analyzes posts on social media as well as police radio chatter and feeds of the local airspace in virtually any region. To detect rumblings of unrest and alert the public. At the moment, the tool has its gaze trained on Baton Rouge, where protesters backed by the New Black Panther Party have gathered for a rally. "I'm looking for any indication they are coordinating skirmishes. Using IBM's Watson AI, the tool not only examines large collections of tweets but -- somewhat eerily -- also can go through a single user's timeline and, with Watson's machine learning technology, offer an analysis of that user's "trustworthiness, propensity toward violence [and] openness," the Jester said. That information, he said, could hold clues to a criminal's intentions. If the Jester's name sounds familiar, that's because the hacker has appeared elsewhere -- on Time's list of most influential Internet personalities, on CNN and, according to a recent blog post, on an upcoming episode of USA's "Mr.

As the country reels from the spasm of gun violence that killed two black men and five police officers this week, a prominent digital vigilante is using an online tool he hacked together to keep an eye on hotspots that seem at risk of boiling over into bloodshed. The Web app, which is powered partly by artificial intelligence, analyzes posts on social media as well as police radio chatter and feeds of the local airspace in virtually any region. To detect rumblings of unrest and alert the public. At the moment, the tool has its gaze trained on Baton Rouge, where protesters backed by the New Black Power Party have gathered for a rally. "I'm looking for any indication they are coordinating skirmishes … I guess I'm expecting trouble in that location, so [I] have it trained on Baton Rouge preemptively," said the creator of the site -- who goes solely by his Internet pseudonym, the Jester -- in an interview with The Washington Post.